AI News Digest 2026-10-08
台本で使った記事
特集
中堅コーナー
開発者コーナー
AIツール紹介コーナー
速報コーナー
参考記事一覧
参考記事一覧を表示
- Claude Haiku 5.5 arrives with massive price cuts proving the AI pricing arms race is far from over
- Introducing Mistral Large 4: Le chonk
- EmbeddingGemma 2
- Google's new image model Nano Banana 2.1 generates better images for less money
- OpenAI launches Decisions API that reduces complex evaluations to yes, no, or pick one
- OpenAI dumps 372 AI-generated math proofs on GitHub, telling the academic world to keep up
- Google says 180 billion images and videos now carry SynthID watermarks as detector goes public
- OpenAI、EUのChatGPTとCodex出力に不可視の透かしを導入
- ChatGPT rated "unacceptable risk" for teens after parental alerts failed during suicide conversations
- Muse launches on the iPad
- Elon Musk Clarifies TSMC's Terafab Role; Grok Adds Rival Models - Investor's Business Daily
- Elon Musk’s xAI Blocks Minnesota’s Landmark Ban on AI Nudification Apps - PetaPixel
- Grok AI Says ‘Jesus is Not a Fake God,’ Cites Evidence For Christ and Resurrection - Beliefnet
- Tencent Music integrated DeepSeek into its streaming service. Now parent company Tencent is among the biggest backers of the AI company’s $12B+ funding round (report) - Music Business Worldwide
- What is Beam? The buzzy American AI model being called the 'DeepSeek of the West' - Business Insider
- 16 AI Models In One Month: DeepSeek, Chinese AI Firms Step Up Launch Race Despite Anthropic Warning - NDTV Profit
- DeepSeek, Moonshot AI Plan 2025 IPOs After Investment Surge - 조선일보
- GPT-6 and Intelligent UI for everyone
- Anthropic gives more security teams access to Claude with fewer safety restrictions
- OpenAI agents tried to hack Wikipedia tools and flooded it with traffic
- Anthropic is giving startups a free year of Claude Team and $1,000 in credits
- Show HN: Bigwords.page – Turn any screen into a sign. The URL is the app
- GitHub Incident with Git Operations, Pull Requests and Actions
- Visa, Mastercard, Major Banks Facing New Litigation over 'Anticompetitive' Fees
- Animated ASCII Art for Web Pages
- Anti-Patterns in Software Blogging
- Show HN: A walkable 3D art history museum built from Wikipedia
- God of War on PSP, recompiled to WebAssembly and running in the browser
- Shipping JPEG XL in Chrome
- A font recreated from photographs of classic Commodore 64 keycaps
- v2.1.293
- v2.1.292
- v2.1.291
- Remote control for local agents
- v1.18.35
- Helping teens learn, plan, and shape the future of AI
- Radisson Hotel Group brings hotel discovery into ChatGPT
- How Jump Trading is scaling quant research with ChatGPT
- Advancing computer use with Ironclad
- Atlassian and OpenAI expand partnership to turn enterprise knowledge into action
- Introducing Playground: Create and play custom games
- Unlocking Earth AI’s planetary geospatial foundation models for global public health
- Agent Lightning v1.0: A 3,500-Line Lightweight Agentic RL Framework for Training Agents with Real Harnesses
- What AI gets wrong and what failure teaches us
- Multimodal open d1 decision models for the edge
- One Model Family, Two Gold-Level Results: Fine-Tuning Nemotron for IOI and IMO
- Secret protection must scale with software
- Building Git infrastructure for agent-scale development
- Building an evidence-grounded agentic security operations harness on Cloudflare
- The keys to the Internet change on October 11. Are you ready?
- NTS: Authenticated Time at Meta
- Introducing Warm Agents, a TeamCity Plugin
- Faster Developers Don’t Make a Faster Team
- Start Your Next KMP App With Kotlin Toolchain 0.13
- Your Next Action Is Two Dots Away
- Our Framework for Reviewing AI-Generated Code
- Amper Is Now the Kotlin Toolchain: Moving to the Kotlin Blog
- Java and Kotlin by IntelliJ IDEA: The LSP-Based Extension Reaches Release Candidate
- JetBrains Joins the Open Source Security Foundation
- The Machines that Make the Machines
- Scaling Decision Optimization to 100 Million Variables and Beyond with mPDLP in NVIDIA cuOpt
- Faster Scientific Image Analysis with NVIDIA cuPhoton
- How DOCA GPUNetIO Unifies GPU-Initiated Networking Across the NVIDIA Software Stack
- AICR v1.0: Open, stable, and verifiable GPU cluster configuration
- What’s New for Game Developers: DLSS 5 with 3D-Guided Neural Rendering, NVIDIA ACE Updates, and New RTX Kit Capabilities
- Fraudster jailed for using 10K bots and AI songs to outstream Taylor Swift
- ChatGPT is getting a lot more visual, with the launch of a new interface
- Meta rolls out new AI tools to detect ads that secretly lead to child sexual abuse material
- Healthleap raises $38M for its AI that flags hospital patients who may need a closer look
- Tony Fadell on why the first wave of AI gadgets failed — and what comes next
- Google experiments with an AI-powered gaming platform
- OpenAI’s Alexander Embiricos is coming to TechCrunch Disrupt 2026 — days after the launch of Dots
- Get hands-on: The full lineup of interactive roundtables at TechCrunch Disrupt 2026
- 6 days to TechCrunch Disrupt 2026: Save on your pass before doors open
- Ex-Ramp engineers raise $20M for platform Melius after scrapping their first product
- How AI decision models could change content moderation
- AI computing startup Lambda to raise $4B ahead of planned IPO
- The next hurdle for AI agents: getting websites to let them in
- Hark releases an AI personal assistant with a focus on privacy
- Mirror Particle is building a ‘world model’ of human behavior
- LibreOffice says ‘no AI’ is now a software feature
- Everything announced at Microsoft’s Surface Laptop Ultra event
- Microsoft is giving Copilot more control over Windows and your files
- Surface RTX Spark Dev Box is available for preorder for $5,999
- ChatGPT is getting college planning tools
- Google invests millions in Mark Zuckerberg’s efforts to create a ‘virtual cell’
- AI could upend food delivery
- We can’t just change the definition of ‘recording’
- Survey Finds AI-Generated Code Increases Debugging and Failure Rates and Creates a Comprehension Gap
- AI in Production: What Breaks, What Works, and Who Approves It? | InfoQ Webinar
- Article: Building a Session-Ordered Kafka Pipeline in Go
- Cloudflare Uses an AI Harness to Probe and Harden Its WAF
- Spanner Omni Reaches GA, Replacing Google's Atomic Clocks and File System with Software
- QCon London 2027 Announces 15 Tracks on Production AI, Architecture, and Engineering at Scale
- HiPHI: A Large-Scale Benchmark for High-Precision Human Motion and Object Interaction
- Zuckerberg's Biohub leads a $1.8 billion push to build AI models that predict cell behavior
- Google bets Gemini can turn casual players into game developers with new Playground feature
- Meta、Muse向けESP32とLinuxのSDKをApache 2.0で公開
- Amazon EKSがKubernetes 1.37に対応 ゼロスケールが既定で有効
- Effect 4.0、依存ゼロのコアで公開 最小バンドルは7.1kB
- Handlebars 4.7.9以前、オブジェクト渡しのcompileにコード注入
- Next.js 16.4、Cache Componentsを全アプリへ推奨 新規は標準有効
- ミスターマックスのアプリとECに不正アクセス、最大173万人分の会員情報が流出
- 楽天ドライブに不正アクセス、管理用アカウント経由で1万5382件の保存データが閲覧
- OpenAI、CodexのAuto-reviewをChatGPTアカウントの全ユーザーに無料開放 利用枠も消費せず
- Claude、Google Docs・Sheets・Slidesのサイドバーで直接編集可能に 有料プランで公開ベータ
- npmの@subql/common 5.8.3に認証情報収集のコード、導入とimportの両方で起動
- Dell、Kubernetes接続ストレージの認証欠如を修正 最大CVSS 10
- BuildKit、共有キャッシュ経由で別イメージの層をすり替える欠陥を修正
- Nuxt 4.6、サーバーAPIをNitroから分離しCLIの起動を高速化
- AWS、音声エージェント向けNova 2.5 Sonicを東京を含む4リージョンで提供
- Reflection、5010億パラメータのオープンモデルBeamを今月公開へ
- Meta AI Open-Sources Rebalancer: A C++ Assignment Solver That Runs About 40 Million Placement Problems a Day
- A Developer’s Guide to Laya: Zero-Shot Decisions and Calibration
- Reka Releases Rho-1: A 19B Omni-Reasoning Model That Understands, Generates Video and Outputs Robot Actions in One
- Beyond Domain-Specific World Models: JEPA-Anything Uses 1 Recipe for 7 Fields
- Training Numerical Intelligence via Auto-Diagnosis and Skill Discovery
- REACT: Physically and Chemically Consistent Reconstruction of Marine Active Tracers
- Proxy Confidence: Auditing Black-Box LLM Agents with a Surrogate's Log-Probabilities
- SGAnalog: An End-to-End Circuit Benchmark from Open-Source Silicon Tapeouts
- MLLMs Fail to Refuse when Using Tools Agentically
- Retrieval-Augmented Large Language Model Decision-Making for Autonomous Driving Guided by Chinese Philosophical Wisdom
- LatentQuant: Preserving the Policy-Facing Latent Contract under NVFP4 VAE Quantization
- ROAR: Unifying Runs across Heterogeneous AI-Driven Research Systems
- Teaching Agents to Code Reliably
- Behavioral History Outperforms Descriptions of the Person for LLM Synthetic Personas
- SkillScriptBench: Benchmarking Self-Evolution of Executable Agent Skill Packages Beyond Markdown
- Exploration-Preserving Policy Optimization
- Beyond the Parameter Monolith: Reconstructive Memories, Executable Skills, and Residual Assembly for Language Models
- A Quantitative Analysis of Graph Representation Strategies for Cyber Attack Detection
- The Reported Engagement with AI Level (REAL) Rating: A Framework for Disclosing Human-AI Collaboration
- Agent Policy-Value Audit: Separating Transition Composition from Event Selection in Financial LLM Agents
- The Cost of a Hop: Benchmarking NLIP and A2A
- Discrete Diffusion for Large Graph Generation via Structural Candidate Restriction
- Reinforcement Learning with Comparative Evidence for Social Intelligence
- Self-Propagating Misalignment in LLM Agents, and Why Auditing or Disabling Memory Is Not Enough
- Towards Safer Autonomous Driving in an Open World: A Dual-Process Approach
- The Independence Prior of SAEs Fragments Visual Concepts
- CUAWright: A Minimal Unified Interface for Digital Agents
- InvestigationWorlds: An Agentic Environment for Legal Investigation
- Auditing Pairwise Equivalence Judgments: Self-Critique Effects and Diversity Measurement in Multi-Agent Hypothesis Generation
- Agentic Cognitive Depth: Operational Criteria for Evaluating LLM Agents
- SHarP: Saliency-based Pruning of Agent Harnesses
- Language Model Activations Inhabit Privileged Error-Correcting Basins
- EvalResearchBench: Can AI Agents Design Their Own Evaluations?
- Agentic AI with Structured CoT for Enhancing AI's Spatial Intelligence: Visualization and Reasoning of Rotation
- MemLeak: Cross-User Semantic Leakage in Multi-Tenant AI Agent Memory
- Asynchronous Is Nearly Free for Evolution Strategies on Long-Horizon Agentic Tasks
- ALoDLM: Adaptively Looped Diffusion Language Models
- Fine-Tuning VLM for Enhancing AI's Spatial Intelligence: Understanding 3D and 2D Rotations
- TCMClinicalReason-Bench: Can Language Models Reason from Pathogenesis to Prescription over Real-World Clinical Cases?
- On the Steering Dimensionality of Refusal in Language Models
- Spec2Game: Can LLMs Generate Complete Playable Games from Detailed Specifications?
- CADForge: Agentic Single-View CAD Reconstruction with Explicit Geometry Reasoning
- Dense Neuro-Symbolic Reasoning in a Unified Geometry State
- VIGIL: Verifier-Informed Gated Improvement Loop for Spreadsheet Question Answering
- LMBuild: Evaluating LLM Agents for Generating Buildable and Functional Structures
- Language-Conditioned Token and Reasoning Efficiency in Large Language Models: A Paired Cross-Lingual Study Protocol
- EnvDreamer: Large-Scale Multimodal-to-Environment Generation for Embodied AI
- MOIRA: Mass-Oriented Indexing with Ragged Attention for Long-Context Decoding
- From Latent Space to Jacobian Space: Measuring, Evading, and Training Against Safety-Content Accessibility
- Bidirectional Preference Synthesis: Learning Prompt-Conditioned Preferences from Boundary Failures
- Suppressing Pressure, Amplifying Evidence: Self-Guided Attention Steering to Mitigate Sycophancy and Stubbornness
- ShadowMiner v1 - An Experience Report on Implementing and Measuring a Problem-and-Hypothesis Discovery Engine
- Functionally Equivalent or Not? Graph-Grounded Differential Surrogate Execution for Code Equivalence
- Do Tool Calls Execute as Intended? Measuring and Repairing Intent-Execution Correspondence in LLM Agents
- AgentPersonaBench: Benchmarking Persona-Driven User Simulation
- TrustMed-RL: Long-Horizon Reinforcement Learning for Evidence-Grounded Clinical Diagnosis
- TimeNet: An Extensible Unified Data Infrastructure for Next-Generation Temporal Foundation Models
- CORE-RL: Confidence-Oriented Reliability Evaluation of Black-Box Reinforcement Learning Policies
- Asking Earns Nothing: Scoring the Decision to Act in BFCL Multi-Turn
- MERCI Cards: An LLM Evaluation and Deployment Framework for High-Stakes Domains
- What Does a Harness Buy? Tokens, Mostly
- CRAFT: An Agentic Spreadsheet Form Filling System with Template Awareness
- RAGrasp: Geometry-Semantic Template Retrieval and Grasp Transfer
- Causally Fair Generation with Large Language Models
- Understanding Generative AI Use in Programming MOOCs: The Role of Course Context and Learner Characteristics
- Trinity: Self-Evolving Vision-Language Models with a Self-Verifier
- Reactivating Alignment: Defending LLMs from Jailbreaks via Intention-Aware Input-Output Matching
- Semantic Causal-Factor Inference from Aviation Incident Narratives Using A Variational Autoencoder with Cosine-Similarity-Based Reconstruction
- InferOpt: Constrained Multi-Objective Search for LLM Inference Configurations
- VCLMU: Mechanism-Centric Virtual Cell World Modeling for Perturbation Response
- ManifoldCache: Training-Free Diffusion Acceleration via Constraint Manifold Caching
- Towards Credible Agent-Based Policy Simulations: Disentangling Opportunities and Preferences in a Financial Inclusion Case Study of Egypt
- EvoCast: Reliable Autonomous Research Agents for Iterative Forecasting Architecture Evolution
- Action-Consequence Alignment for Reliable Planning and Self-Improving in Latent World Models
- Autonomous Structuring of Radiology Reports Across Modalities at Archive Scale Using an Open-Weight Large Language Model
- Decide, Ask, or Defer: Clinical LLMs under Incomplete Evidence
- Label Agreement Does Not Measure Authorization
- $\mathrm{TRIZ}^{a}$: Guiding Agent Evolution from Pattern Recognition to Solution Invention
- Recursive Improvement of a Differentiable Scientific Software Ecosystem
- Strong Helps Weak: Directional Cross-Modal Alignment Transfer in Multi-modal LLMs
- Bounds, Decompositions and Null Behaviour of KRATOS: A Mathematical Specification of a Recognition-Comparability Diagnostic
- LatentIndex: Cross-Layer Sharing with Layer-Specific Selection for Sparse Attention
- SEIS: Self-Evolving Inference Systems
- PermVLA: Factorization Order as a Regularizer for VLA Learning
- Efficient Neural Surrogates for Linear Radiation Transport on the Lattice and Hohlraum benchmarks
- A Tropical Geometry View of Forgetting: A Per-Unit Projector for Knowledge-Preserving Fine-Tuning
- MASBench: Benchmarking LLM-based Multi-Agent Collaboration under Partial Observability
- When Debate Helps: Proposal Supply and Verification-Aware Readout in Multi-Agent Reasoning
- RAGStress: A controlled benchmark for evaluating retrieval-augmented generation under knowledge-base degradation
- Penumbra: Sample-Efficient Adversarial Search for Regulatory Obligations
- Not Self-Decidable: LLMs Cannot Draw the Boundary of What an Agent Verifier Can Check
- Learning to Clarify Underspecified Intents Under Limited Interaction
- PyINE: A Framework for Scalable Elicitation and Oversight via Code Execution
- Toward a Locally Deployable Agentic Co-Scientist: Small-Model Planning for Early-Stage Drug Discovery
- Formalizing the Moral Evaluation of Speech Acts: Truthfulness, Lies and Ethical Dilemmas
- Agentic discovery of blood biomarker from distilled private health records
- Dynamic Routing as a New Dimension for Test-time Versatility of LLMs
- When Is Enough Enough in Self-Evolving LLM Systems?
- Not All Answers Are Contextually Persuadable: Inference Dynamics in Large Language Models under Contextual Influence
- Do More Modalities Always Help? A Geometric Perspective on Missing-Modality Robustness
- Pressure, Context, and Machine Self-Control: A Criminological Test of Reward Hacking in Generative AI Models
- Knowing the Store: What a Memory Backend Must Write Down Before an Agent Can Read It
- MedImageOSWorld: Benchmarking GUI Agents for Medical Image Consoles
- Agent Behavior as Code: Efficient and Robust LLM Agents with Programmatic Specifications
- DICE: Decoupling Capability from Intervention Necessity in LLM Tutoring
- MemTrace: State-Consistent Memory for Long-Horizon Coding Agents
- GitSwarm: Decentralized Compounding Inference
- From Memory to Guide: Spatio-Temporal Composer for Procedural Coding Memory
- SpecFold: Folding Multi-Branch Redundancy for Faster Speculative Decoding in Diffusion Language Models
- ForkPilot: Self-Evolving Policy for Retrospective Search in Long-Horizon Agents
- Self-Evaluating Recursive Agents
- VideoResearchAgent: Grounded Task Synthesis and Sim-to-Real RL for Open-Web Video Research
- Monitorability Disposition in Large Reasoning Models
- Are We Measuring Scientific Intelligence? Rethinking the Evaluation of AI Scientists
- Zero-Shot Time-Series Question Answering via Decoupled Perception and Reasoning
- BACAM: Behavior-Aware Continual Agent Merging for Multi-Turn Interaction
- Runtime Authorization of Self-Generated Subgoals in Long-Horizon Tool-Using AI Agents
- Do LLMs Understand Sequential Structure? A Controlled Study of Inference and Generation
- CIPO: Counterfactual Imagination Policy Optimization for Adaptive Tool Granularity Selection
- EmoRSS: Mitigating Emotion-Induced Over-Refusal in Large Language Models
- How corner is a corner case? Percentile control for highway scenario generation
- EVISKILL: Grounding Skill Evolution in Replayable Evidence
- CI-JEPA: A Counterfactual Analysis of Latent Representations in Joint-Embedding Predictive Architectures for Self-Supervised Learning
- Why, Where, How: Taxonomy-guided Error Grounding for Code Repair in NL2SQL
- SharpDraft: Accelerating Long-Context Speculative Decoding with Cardinality-Aware Query Scaling
- LexiHorizon: Stabilizing Reinforcement Learning for Long-Horizon Deep Search
- Memory Canonicalization: A Framework and Benchmark for Cross-Model Drift in Persistent LLM Memory
- AutoSciBench: Autonomous Benchmark Generation for Evaluating Scientific Agents
- R1A-PC: Physics-Guided Electromagnetic Inversion of Three-Dimensional Human Point Clouds in Complex Static Environments
- Beyond Instruction Following: Learning Grounded Skill-Following with Skill Contracts
- Memadapter: Counterfactual Adaptation Against Memory-induced Sycophancy
- A Safe Action Is Not Enough: Feasible-Future Decoding for Vision-Language-Action Policies
- CreativeFlow: A One-to-Many Analogical Relation Transfer Method for 3D Asset Generation
- AECG: Asymmetric Experience Consolidation and Governance In Multi-Agent Systems
- From Scientific Observations to Mechanisms: Benchmarking Hypothesis Generation by AI Scientists
- Image Synthesis as an Intermediate for Controllable Time Series Generation
- Safe Context Switching for Agents in the Wild: Mitigating Subspace Interference via Orthogonal Adaptation
- Look Before You Leap: Thermodynamic Arbitration of Parametric and Non-Parametric Knowledge in LLM Agents via Self-Regulating Memory Architectures
- GFGE: Unifying Explainable AI Methods through an Interpretation Framework
- Learning from imperfect teachers for low-resource acoustic generalization
- When Agent Context Goes Stale: Incoherence in Volatile Agent Context
- Fusion is the New Mutation: Bandit-Guided Evolution on Workflow Graphs
- Readable Before Actionable: Causal Tracing of Indirect Prompt Injection
- MESH-Harness: Self-Improving Agent Harnesses via Bandit-Guided Compositional Evolution
- AgentDiscover: Autonomous Discovery with Minimal Search Scaffolding
- EnGRICH: Enhancing Generative Reward Modeling with Critiques from Humans
- Learning Field Reconstruction from Incomplete Data by Globally Correcting Local Estimates
- Sibyl: An Efficient Small-large Model Collaboration Framework for Long-horizon Tasks
- MMPostTrainBench: Benchmarking Autonomous Research for Multimodal Post-Training
- CASE: Cost-Aware Stopping for Efficient Long-Video Agents
- PharmAgent: Constraint-Aware Search with Frozen Language Models for Molecular Optimization
- TeleTune: Evolving Agent Skills From Offline Telemetry
- AI Safety via Debate is Compromised by Cognitive Biases
- Hallucination Across the Reasoning Lifecycle: Interface Visibility, Causal Evidence, and Release Control in Large Reasoning Models
- Hierarchical Reinforcement Learning with Stable Temporal Abstraction for Language Model Agents
- CodeForge-MA: Execution-Verified Multi-Agent Learning with Language-Conditioned LoRA for Multilingual Code Generation
- SkillGATE: Gate-Aware Monte Carlo Tree Search for Skill Retrieval
- The Functional Structure of Post-Compression Recovery in Low-Rank LLMs
- What Does Fr\'echet Distance Measure? A Directional Decomposition
- Have I Scene This Before? Spatially Grounded Conversational Memory for Complex Queries in Egocentric Assistants
- DelegationBench: Measuring When AI Agents Should Ask Before Acting
- More Claims, Less Evidence: Bounded Verification of AI-Generated Digital Knowledge Artifacts
- LifeLong Digital Twin: A Unified Modeling Paradigm and Agent Harness for Event-Driven Lifelong Health State Trajectories
- Factoriax: A GPU-Accelerated Factorio-Style Simulator for Reinforcement Learning
- Better Retrieval, Limited Clustering Gains: A Controlled Study of Multilingual Company Entity Resolution
- DREAM: Dynamic Resolution Assignment For Multimodal Multi-agent Debate
- A Framework for Automated Multi-Source Satellite Data Analytics and LLM-Based Report Generation
- The GenAI4IDN Benchmark 3.0 - a Public Tool to Assess Generative AI Tools for the Design of Interactive Digital Narratives
- Visual Grounding Safety in Vision-Language Models
- How Should a Prompt Optimizer Spend a Tight Budget? BudgetAPO with Noise-Adaptive Evaluation
- Do Time-Series QA Systems Read the Time Series? Evidence Use and Reasoning Reliability
- Toward AI Trustworthiness: Finding Analytically Proven Forward-Invariant Sets for AI-Controlled Systems
- From Token-Max to Outcome-Max: How You Use AI Determines Its Productivity
- Second-Order Problem Solving for Recursive Self-Improvement in Formal Verification
- FreSia: Frequency-Semantic Instantiation and Alignment for Multivariate Time Series Analysis
- Mining Agent Skills from Production Traces
- Agentic-ZTA: A Multi-Agent Architecture for Autonomous Zero Trust Enforcement
- Topology-Conditioned Backdoors: Language Models That Insert Vulnerabilities When They Infer They Are in a Multi-Agent System
- A Testable Theory of Atomic Features
- DiMOS: Doob-Guided Inference-Time Multi-Objective Search for Scientific Design
- Nash Equilibrium Text: A Game-Theoretic Decoding Framework for Text Generation
- Data-Driven Personas for Survey Simulation: Insights into Simulation Alignment Across Data-Access Regimes
- Measurement-First Auditing of Agentic Leaderboards: Contamination Susceptibility, Matched-Control Re-evaluation, and Scorer Validation
- Request Order Matters: Cache-History Sensitivity in Selective KV-Cache Reuse for Rolling Agents
- MiniCorp: The Last Mile of the AI Agent Firm
- VERA: Scaling Verifiable Environments for Agentic co-Evolution
- Process Constitutions and Process Stewards: Towards the Next Generation of BPM for Agentic Organizations
- OntoInk: Interactive Ontology Visualization, Validation, and Reasoning
- Grounded Joint-Attention Other-Play for Zero-Shot Coordination
- RocketAgent: A Long-Horizon Engineering Agent for Multidisciplinary Design of Liquid-Rocket Thrust Chambers
- Benchmarking Jailbreak Guardrails for Embodied Agents
- On Impact of Loss Function on the Performance of Neural Networks in Melanoma Diagnosis
- Impact of Data Augmentation on Confidence Calibration in Melanoma Classification
- MS-Exam-Gen: Source-Grounded Benchmark Construction for Evaluating LLMs on Textual Multiple Sclerosis MRI Knowledge
- Auditable Clinical Timeline Reconstruction with Provenance-Aware Evidence Graphs
- Bridging the Evidence-to-Execution Gap:A Reflective Agent for Multi-Objective Peptide Design
- Judged Useless, Queried Anyway: Tool-Using Agents Rarely Turn Their Own Evidence Judgments into Stopping Decisions
- Copies or Sources? Measuring How LLM Aggregators Count Restated Evidence in Multi-Agent Systems
- Do Small Language Models Learn to Negotiate? A Controlled Scaling Study of RL-Trained Sellers
- AI-Decision Checkpoints for AI-Augmented Business Process Management: Framework and Educational Instantiation
- From Papers to Mechanisms: An Evidence-Grounded Knowledge Substrate for Scientific Language Models
- Evolving in Thought Space: Training a Small Model at Test Time Unlocks Better Discoveries
- DPNL: A DPLL-based Algorithm for Probabilistic Neurosymbolic Learning
- Teaching a Minimalist Machine to Discover Recursive Programs for Arithmetic
- RollPlace: Improving Macro Placement via Monte Carlo Rollout Search
- CRAFTER: Causality-based Self-adaptation for Autonomous IoT Systems
- ImproveAnyTask: An Autonomous Post-Training Harness for Iterative Model Self-Improvement
- GraphDecide: Benchmarking System One Models on Graph Tasks
- Capability-Driven Self-Evolution of Agent Memory
- CVIF: A Criticality-Driven Visual Intervention Framework for Geometric Diagram Understanding in MLLMs
- What Did the Agent Actually Do? Evidence-Grounded Oversight for Long-Horizon Agents
- From Benchmark to Bench: Can Agents Survive Real-World Drug Discovery?
- Multimodal Safety Evaluation Should Measure Controllability Beyond Classification
- Normality Constraint Learning: Adapting Foundation Models for Time Series Anomaly Detection
- AgentPrivArena: Evaluating and Auditing Real-world AI Agent Privacy
- GPlaceRL: An Open-Source Graph Reinforcement Learning Framework for Detailed Placement
- polyview: A Python package for multi-view machine learning
- You Changed Your Mind, The Model Didn't: Demystifying Intent in Multi-Turn Dialogue
- ANT: A Multi-Granularity Network Traffic Dataset and Benchmark for Agents Behavior Auditing
- Proof-Grounded Patient-Specific Clinical Explanations from Knowledge-Graph Reasoning
- Signature-Based Feature Learning for Human Activity Recognition: A Reproducible Machine Learning Study of Representation, Depth, and Model Choice
- HERA: Harness-Environment Co-Evolution for Reliable Agentic Abstention
- DGA-Muon: Decoupled Geometry-Aligned Adaptive Scaling for Muon
- Mind the Execution Gap: Action-Semantic Mismatch in World-Model Control
- The Review Lottery: Calibrating an Observational Estimator of Peer-Review Noise (ICLR 2017-2025)
- Can Agent Harnesses and Inference Engines Hear Each Other? The HEAR Protocol for Agentic LLM Serving
- FREA: A Multi-Source Expert Benchmark for Reaction Feasibility Verification
- Collective intelligence through aggregation
- Language models can notice an impossible engineering problem yet still report it as solved
- The Pushback Paradox: A Two-Probe Diagnostic for Language Model Compliance
- Conditional Rank Allocation for Taxonomy-Aware Medical Language Model Adaptation
- Back to the Future: Rethinking EDA Infrastructure for Agentic Systems in Chip Design Verification
- TasteVal: Measuring the Experimental Research Taste of AI Systems Against Human Experts
- BiasFlow: Geometric Monitoring and Backbone Regularization for Spurious Feature Reliance
- Orchestrating Specialized Agents for Trustworthy Enterprise RAG
- AI Agent Pull Requests on GitHub: Frequency, Structure, and Merge Conflict Rates
- Using Process Mining to Generate AI Agents from Software Engineering Process Records
- Beyond Private Training: The New Landscape of AI Privacy
- Response Variability and Stability in Human Reasoning
- Efficient Analog-Initialized Latent Transport for Spatially Coherent Probabilistic Downscaling from Global to Kilometer Scales
- Logit-Aware MIMO AirComp for Distributed Mixture-of-Experts LLM Inference over Wireless Edge Networks
- LoRA Adaptation Strength in Aurora-WRF: Trade-offs in Regional Heavy-Rainfall Forecasting
- Global Evaluation of AI and NWP Precipitation Forecasts During Atmospheric River Events
- BridgeCast: Bridging Ocean Wave Forecasts to Reanalysis via Flow Matching with Exogenous Variables
- Do Motion Tokenizers for Co-Speech Gesture Generation Encode Gesture Semantics?
- OceanMind: A multi-agent AI system for ocean diagnosis
- ProsaBuddy: Assisting Mechanized Real-Time Schedulability Analysis with LLM-based Agents
- WAMJET: A Harness for World Action Model Acceleration
- From Requirements to Attack Trees: Grounded LLM Agents for Design-Time Security Review
- The Score Is Not the Structure: Brain Alignment and Cross-Lingual Transfer
- Distribution Matching Evolutionary Algorithms for Rare Event Sampling
- SYNLAT: Syntax-Aligned Text-Latent Compression for Chain-of-Thought Reasoning
- Beware EviLLM: Enabling Vulnerability Injection via Large Language Models
- Retrieval-Centric Deep Learning in Growing Nonparametric Neural Networks
- GOTT: Object-centric Dexterous Manipulation with a Reusable Cross-Embodiment Primitive
- Streaming Multi-Track Timeline Control for 3D Human Motion Generation
- A Step Towards Forgetting: Optimiser History and the Loss of Answer Mass
- An Executable Benchmark for LLM-Based HLS Repair:Design Complexity and Repair Underconstraint
- Solving VeriContest with a Lean-Backed Rust Verifier
- Selective Backpropagation for Efficient Few-Shot Class-Incremental Learning
- Slaying the Hydra: Interaction-Aware Circuit Discovery in Language Models
- MAGEFormer: Learning Metric-Consistent Representations for Anisotropic CT Segmentation
- Reward-DAgger: Robot-Gated Interactive Imitation Learning with General-Purpose Progress-Based Reward Models
- A study on human-agent teaming through spoken interaction: the impact of human individual traits and agent characteristics
- AEGIS: Differentiable Mars Climate Model with Neural Closures
- Dependable AI-Assisted Engineering: A Formal Framework for AI Participation and Assurance in Safety-Critical Workflows
- DUET: Co-Evolving Solver and Grader Agents
- Robust blind unmixing: A geometric approach to overcoming basis variation
- SEER: Self-Evolving Event Reasoning and Retrieval for Time Series Forecasting
- Agent Reliability Profiles in Financial Services
- Representational Control over Self-Report & Behavior Coherence in LLM Risk-Taking
- PB-GRPO: Learning Socially Adaptive LLM Agents from Persona-Driven Simulation with Preference-Batched GRPO
- Agentic Resource Allocation for Batch Multi-Objective Bayesian Optimization in Autonomous Materials Discovery
- Dual-Scale Relational Graph Transformers for Ecosystem-Aware Fraud Detection
- Trajectory-Derived Confidence for Reliable, Resource-Aware Clinical Text-to-SQL Agents
- How RL Reshapes LLM Reasoning: Transferability, Coverage, and Scaling Laws
- Language Model Fingerprinting Requires Rethinking Watermark Teachers
- Risk-Calibrated Proposal Transport for Finite-Particle Diffusion Steering
- Clean: Second-order LLM Training at Linear Memory Cost via Nystr\"om Sketching
- CurveCodec 2: Skeleton-agnostic animation compression with a learned entropy model
- FTD-GNO: Memory-Efficient Graph Neural Operators through Functional Tensor Decomposition of the Kernel
- Adaptive Operator Selection in Bilevel Large Neighborhood Search for Electric Autonomous Dial-a-Ride Problem under Uncertainty
- Attention-Based Surface Representation Learning for Robot State Prediction and Open-Ended Surface Classification
- Benchmarking Psychological Dynamics in Generative Agents
- Playing social deduction games with reinforcement fine-tuned large language models
- AI-Enabled Quality Assurance for Multiple-Choice Assessment Items
- Rethinking Self-Distillation for Multi-Teacher Capability Merging
- First-Order Steering: Translating Weight Adaptation into Activation Steering
- CyTReX: Explainable AI-Based Cybersecurity Threat Reasoning Framework for DER Networks
- Scaling of Wireless Foundation Models via Representation Diversity and Multi-Branch Architectures
- RRM-GPT: A Framework and Vision for Radio Resource Management Foundation Models
- Questioning the Questions: Sustaining Self-Evolution in Reasoning Models
- What to Preserve in Recursive Computation: A Local Predictive Sufficiency Principle
- GrayShield: Bit-Level Sanitization for Transformer Model Supply-Chain Security
- System One Models for Wireless Decision-Making:Applications and Performance Evaluation
- A Bird's-Eye View of Iterative Reward Design
- Human Behavior-Informed Crash Scenario Generation with Real-World Crash Priors for Autonomous Vehicle Safety Evaluation
- COPEX: Benchmarking LLM Robustness to Adversarial Context Across Model Context Protocol Layers
- GlitchPatch: Repairing Glitch Tokens in Frozen Language Models via Local Retokenization
- XTurnix: Large-Scale Self-Supervised Turn Control through Two-State Binary Decisions
- Understanding and Mitigating Hallucination Escape in Tool-Using LLM Agents
- Specific Algorithmic Interpretability of Neural Networks: A Case Study on Textures
- Maximizing $p$-Mean Social Welfare in the High-Multiplicity Setting: Few Agent Types and Few Item Types
- MaDeL: Manifold-Decomposed Feature Losses for Generative Modeling
- AgroGround: Multi-Granularity Grounded Recognition in Agriculture
- COSMOS: Soft Mechanism Mixtures with Verifiable Routing for Long-Horizon PDE Forecasting
- Video2World: Benchmarking Coding Agents for Interactive World Modeling from Embodied Videos
- Guess My Weight: Profiled Side-Channel Recovery of Floating-Point Neural-Network Weights
- Target-free Latent Safety Alignment
- Can LLM Agents Automate Reinforcement Learning for Text-to-Speech?
- Understanding Clustering in Slot Attention via Particle Dynamics
- Localization Lens for Improving Medical Vision-Language Models
- EgoExo-Next:Benchmarking Vision-Language Models on Visual-Option Next-State and Cross-View Reasoning
- Diffusion-Based Stress Testing of Overload Monitoring for Resilient Emergency Cellular Networks Using Internet CDR Proxies
- Quantifying Collusion Among Autonomous LLM Agents: A Statistical Analysis of the Collusion Wiki Incident
- Weight Decay and Neuron Condensation: A Three-Stage Analysis of Two-Layer ReLU Networks
- Cross-Modal Solar Image Synthesis: Adapting the Surya Foundation Model from He I 10830 {\AA} to EUV Translation and Coronal Hole Segmentation
- Coupling Noisy Pairwise Knowledge to the DAG Posterior for Causal Discovery
- Only Project Once: Projection-Adaptive Loss for Exact Constraint Satisfaction
- From Probe Scores to Alarm Policies: Operational Validity of Activation Monitors for Language-Model Agents
- Variational Quantum Attention for Molecular Graph Learning
- ConEx: Human-Interpretable Saliency Maps via Concept-Aware Attribution
- Policy as Data: Replay-Based Policy Dual Averaging via Advantage Regression
- GS-Codec: A Gaussian-Splatting Bottleneck for Neural Audio Coding
- RETRACE: From Entangled Repair Histories to Reusable Experience for CI Repair
- Extracting Persona Subspaces Through Iterative Nullspace Projection For Modulation
- Neurodiversity-Aware Multimodal Affective Computing for Neurodevelopmental Assessment: From Norm-Referenced Classification to Context-Sensitive Decision Support
- VoCa: Designing Speech-Canvas Interaction for Voice-Based Conversational Agents
- Organising Trajectory Evidence for Language-Model Agent Assurance: Fragments, Methods, and the Residual
- Towards Automatically Pruning Logging Code with Coding Agents: How Far Are We?
- Knossos and Ariadne: Benchmarking and Learning Complete Diagram Topology Extraction with Vision-Language Models
- Robot Learning with Visual Predicted Force
- More Value per Key: Asymmetric Sparse Attention for Faster LLM Decoding
- EasyClassifier: Honest, Reproducible Machine-Learning Classification for Researchers Who Do Not Program
- ARISE: Adaptive Agentic Reasoning with Image-grounded Self-Evaluation for Interpretable IBD Assessment
- Active-DiNTS: Active Differentiable Network Topology Search
- Separating Decision Time from Decision Quality in the Real-Time Gap of Distilled Deciders: Evidence from a Game and a Conveyor Simulator
- Viva La Vida: Verification and Accumulation Failures in Multi-Agent Proof Search
- Cluster Validation Indices as Self-Supervised Objectives for Text Representation Learning
- PIT-GCL: Protein Interaction using Topological Graph Contrastive Learning
- Invisible Ink, Visible Lies: How Production Watermarking Causes LLMs to Hallucinate
- Enhancing Long-Video VLM Embeddings with Query-Aware Streaming Latent Reasoning
- Greedy Local Learning for Language Model Pretraining: Gaps and Objective Design
- ResOPD: Tail Residualization for Sparse On-Policy Distillation
- SFlexRCA: Lightweight, Scalable, and Flexible Root Cause Analysis for IIoT Edge Systems
- A Systematic Analysis of the Predictive Power of LM Surprisal in Reading Chinese
- ScopeSAE: Model-Scope Feature Discovery with Interpretable Layer Selection
- Scaling Verifiable Environments for Long-horizon Work Agents
- Why Subliminal Learning Needs So Much Data: A Noisy Inverse View through Steering Vector Recovery
- PreAct-Nav: Agentic Reasoning Before Action for Urban Navigation
- Complex Agents, Shallow Tests: Demystifying and Enhancing Test Adequacy of Agent Harness in the Wild
- TSAE: Structured Sparse Autoencoders for Interpreting Time-Series Forecasting Models
- Assembling Insights for Agentic Machine Learning Engineering Systems
- DiVeR: Decision-Critical Verifier Learning for VLA Test-Time Scaling
- D-DOIT: Training-free Adaptation of Discrete Diffusion via Doob's h-Transform
- A Graph-Based Inspection and Intervention Tool for Assessing Mechanistic Learning in PINNs
- Software World Models: From Consequence Prediction to Decision Value
- TempoBridge: Source-Conditioned Flow Matching with Optimal Transport Couplings for Single-Cell Population Transitions
- A Unified Dynamics Framework for Reinforcement Learning and Classical Control of a Six-DOF Pipeline-Tracking ROV in NVIDIA Isaac Sim
- EDISCO: Equivariant DIScrete Diffusion for Euclidean Combinatorial Optimization
- Kapture: Capturing Cardiac Dynamics with Koopman-Governed Learning for Efficient Radar-Based Electrocardiogram Recovery
- TeleGen: Improving LLM-Based Web Application Generation via Runtime Telemetry
- Hidden Risks of Jev: An Empirical Study of Security, Privacy, and Dual Use
- Operational Abstractions of Neural Network Concepts via Topological Representations
- SparseCraft: Agentic Hardware-Software Co-Optimization for Sparse Computing
- Communication Shapes Collective Inference in Self-Adapting LLM Societies: Evidence from Mafia
- AraYoungVoices: A Diverse L1/L2 Corpus of Arabic Child and Adolescent Speech
- E$^2$-OPSD: Taming Entropy Overshoot in On-Policy Self-Distillation
- LogSig-SSM: Time-Series Modelling with Multi-Scale Log-Signature Compression for State-Space Models
- Discrete Action Matching: Learning Stochastic Dynamics from Samples via State Graphs
- Self-Generated Feedback Destabilizes Test-Time Training: A Causal Decomposition of Long-Horizon Adaptation
- Tracing a Sparse Emotion-Control Circuit in LLM-Based Text-to-Speech
- Direction-Conditioned Policies for Online Goal-Conditioned Reinforcement Learning
- ReMAP: Restoring the Perceptual Cycle with Reasoning-Time Latent Visual Memory
- VulValidate: Auditing Function-Level Vulnerability Labels with Executable Evidence
- Best-of-$N$ Guidance for Test-time Diffusion Alignment
- Bayesian Entropy-based Reordering for Calibrated Diffusion Language Models
- How Does Geometry Enter Generated Motion?
- Cross-Modal Contrastive Learning for the Retrieval of Immunotherapy-Associated Molecular Signatures from Histopathology
- Blocking at the Boundary: Auditing Long-Horizon Agents against Staged Prompt Injection
- Same Predictions, Different Harms: Causal Auditing of Patient World Models
- Vela: Scaling Vision-Language-Action Models with Adaptive Action Curve Parametrization
- EMG-GPT: Predictive Pretraining on Residual-Quantized EMG Tokens for Hand Pose Estimation
- Settling the Computational Complexity of Max-Min Allocation with Ternary Valuations
- Pythia: Toward Foundation World Models for Multimodal Time Series
- StateWise: Diagnosing and Repairing Persistent Operational State Before Agent Actions
- MGPO: Manifold-Guided Diffusion Alignment for Task-Aware Dataset Distillation
- Task-Aware Joint Pruning and Distillation for Efficient Audio Deepfake Detection
- Who Is Your Agent Serving? Provider-Side Indirect Prompt Injection in Proactive Agents
- When Verifiable Counts Depend on Wording: Auditing Wording Robustness in Instruction Following
- Revisiting Ground-Truth Synthesis from High-Speed Video: Exact Validity Conditions and an Audited Consumer Capture Corpus
- Characterizing Parallelism Strategies in LLM Inference: Fundamental Compute-Communication Trade-offs
- Grammar-Guided Code Watermarking with Green Temperature
- CoDance: Learning Reactive and Compliant Human-Humanoid Interaction from Video
- VAMPS: Visual and Motor Policies from Sampling-Based Planning
- Green-Routed Neural Operators:\\Physics Determines Where the Network Reads
- On Semi-Markov Suboptimality in Hierarchical Reinforcement Learning
- IRSTD-Agent: Agentic Infrared Small Target Detection via Zoom-Guided Interaction Learning
- Reflections and Fragments: Securing LLMs Against Sequential Mosaic Attacks
- Rethinking Tabular Foundation Models On Data Streams
- Optimal Control with Learned Critics under Unmodeled State Dependencies
- AIProver: Agentic Auto-Formalization of Mathematical Research via Certificate-Driven Evolving Harness
- AutoDP-LLM: Automating Data Pre-processing for Intrusion Detection Systems using Large Language Models
- Distributed Subliminal Learning: Replacing Model Updates with Random-Carrier Outputs
- Does Explainability Survive Data Drift?
- GNN-CB: A Graph Neural Network Competition Benchmark for Human and LLM Evaluation
- No Concept Escapes the Audit: Auditing-Aware Unlearning for Verifiable Concept Erasure in Diffusion Models
- BeliefGraph-JEPA: Structured Latent World Models for Action-Conditioned Time Series
- A Strong Baseline for Evaluating Vision Encoders in Multimodal Large Language Models
- Hierarchical Time-aware Bootstrapping for Off-Policy Subgoal Value Learning
- Human-Like Attention? A Psychophysical Comparison of Visual Search in Humans and MLLMs
- Population Scaling or Data Dilution? Dynamics of Local Topology Evolution in Decentralized Learning
- Reflecting on Creative-Boundaries with an AI Co-Doodler
- FLEX-WAM: Flexible Block-Causal World-Action Models for Long-Horizon Imagination and Planning
- What Does a Harness Repair? A Preregistered Study of Visibility, Baseline Adequacy and Evaluation Defects
- LiFT: Loop Flow Transformers
- Don't Judge an LLM Only by Its Activations: Discovering Suppressed Safety Features via Counterfactual Activation Potential
- Soft Strategy Selection for Batch-Mode Active Learning
- Cut Binary Cross Entropy: Efficient Large-Vocabulary Loss and Gradient Kernels for Sequential Recommendation
- G-CARB: Graph-Localized Conformal Agent Risk Budget for Compositional Harm
- Scenario-Based Compositional Statistical Model Checking for Safety Specifications
- Can Prosodic Style Be Inferred from Text Alone? Evidence from Unsupervised Acoustic Clusters
- What Does an Observability Foundation Model Know?
- Expanding LLM Reasoning
- AgentDoxx: Agentic Re-identification of Anonymized Text with Web Search
- When the Cross-Silo Federation Goes Offline: Continual Learning for Site Onboarding with Limited Unlabeled Data
- An LLM-in-the-loop RL Framework for Bioinformatics Feature Selection
- Your Unlearning Gives You Away: Identifying Erased Concepts in Diffusion Models
- Kandinsky 6.0 Video: Foundation Models for Synchronized Video and Audio Generation
- SEA-LM: Egocentric Spatial Audio Understanding for Wearable Microphone Arrays
- UndoBench: Separating Task Competence from Recovery Capability in Tool-Using AI Agents
- Transformer-Based Time-Series Inference of Lindblad Dynamics in Open Quantum Systems
- Knowing the Rules, Applying the Rules: Evaluating Language Models on Traditional Chinese Bazi
- Two-Sample Testing via Path-based Inference
- Nexus: An Execution Fabric for AI Agents Across Cloud, Edge, and Devices
- SimpleMark: Fast Multi-Bit Text Watermarking under f -Divergence Constraints
- Rotated, but How Far? Diagnosing and Improving Object-Rotation Reasoning in VLMs
- HLA-WM: Hybrid Linear Attention for Long-Horizon Video World Models
- CIPHER-MoE: Balancing Efficiency and Routing Fidelity in Trillion-Scale MoE Training
- Beyond Semantic Similarity: Performance and Costs of Agentic Retrieval for Complex Tasks
- Relational Synthesis: Structure-Mediated Concatenative Synthesis for Foley and Retrieval-Augmented Audio Generation
- CLARA: Can AI Assess Developmental Appropriateness in Children's Stories?
- Protocol-Sensitive Evaluation of Log Anomaly Detection: Component Costs and Target-Access Sensitivity on HDFS and BGL
- Level-of-Token Diffusion
- Who Keeps the Gains from Personal AI Assistants? Seller Adaptation and the Unassisted in a Language-Model Market Simulation
- Selecting Long-Horizon Trajectories for Reliable and Efficient Terminal-Agent Training
- CACFG: Curvature-Aware Classifier-Free Guidance and Optimal Control
- Hierarchical Reinforcement Learning for Collision-Free Locomotion of an Underactuated Biped
- Curriculum Brain: Constructing Curriculum Knowledge Graphs as a Substrate for Cognitive Diagnosis
- Imagine to Act: High-Fidelity Data Synthesis via Image Editing World Model for Scalable GUI Agent Training
- CCQ: A Multi-State Child Care Quality Dataset to Support AI for Children's Health Research
- Weave Mamba Fusion: Global Cross-Scale Interaction for Lightweight Face Detection
- Off-Policy Merging Beats On-Policy Self-Distillation for Continual Learning
- OGAM: Connecting Systematic Testing to Runtime Assurance through Object-Grounded Attention Monitoring for VLA Policies
- TasteRoute: Personalized Routing for Video Generation
- On Hyperparameter Tuning on the Test Set
- Non-invasive Seizure Detection Using Wearable Wrist-worn Accelerometry and Deep Learning
- Incentive Alignment in Online Experimentation
- UltraDub: Towards Authentic Dubbing by Unifying Visually-Steered Flow Learning and Trajectory Guidance
- ThunderSyncRL: Lossless Acceleration of Agentic Reinforcement Learning
- ReMem: Streaming Video Understanding With Long Context Retention
- Runaway Reaction: When Benign Skills Compose into Malicious Behavior
- Physics-Informed but Not Physics-Consistent: Error Geometry and Subspace Projection for Neural AC Power Flow
- HuatuoGPT-3: RL-Only Domain Adaptation from Base Models
- JLD: Perceptual Distance Through A Jacobian Lens
- Transfer-Stratified On-Policy Distillation for RL-Improved Reasoning Teachers
- EpicWorldModel: Exploration-driven Planning with Latent World Models
- AgentSpy: Making AI Agent Behavior Observable
- Parallelism or Concession? Concurrency-Aware Procurement Negotiation for Agentic Commerce
- Investigating Query-Insensitive Behavior in Spatio-Temporal Video Grounding
- Adaptive Expert Guidance for Efficient On-Policy Reinforcement Learning
- Scalable Minimal-Change Learning for Controllable Image Editing
- Differentiable Bit-Widths: Co-optimizing Pruning and Quantization via SVD for Ultra-Efficient LLM Compression
- LightMTP: Lightweight Latent Multi-Token Prediction
- MercerFlow: Flow Matching in a Kernel-Induced Latent Space for Probabilistic Forecasting
- Representation Disentanglement for Fair Chest X-Ray Diagnosis
- Backdooring Sparse Autoencoders
- Local2Mesh: Spatially Localized Contour-to-Mesh for Left Ventricular Reconstruction from Sparse 2D Cardiac MRI
- ROT: Rotating Hidden States towards Contextual Vectors for Hallucination Mitigation in LVLMs
- A Comprehensive Objective Evaluation of Modern Text-to-Speech for Turkish Using Speech Quality Assessment Models
- Vision Transformer Ensembles for Panoramic Street Segmentation
- Attention Tax, Handoff Tax: A Stylised Model of When Multi-Agent LLM Systems Help
- Do VLAs Understand and Adapt to the Objects They Handle, or Simply Replay Learned Behaviors?
- Boosting Transferable Adversarial Attacks against Deep Reinforcement Learning
- Adaptive Mean Flow for Responsive Closed-Loop Robot Control
- Cross-Lingual Transferability of Training Data Extraction Attacks to Recover Memorized PII
- On the Geometry of Multimodal Saturation: Riemannian VICReg
- Machine learning for journal entry testing: A type-aware evaluation of anomaly detectors under a review budget
- Where Did the Repair First Go Wrong? Localizing the Origins of Silent Failures in Agentic Vulnerability Repair
- Anosognosia in LLMs: Probing Self-Awareness of Quantized Computational Substrate
- Anlu: Enabling In-Context Time Series Anomaly Detection in Foundation Models via Counterfactual Supervision
- Correct Code, Broken Contributions? SWE-CC: Benchmarking Repository Policy Compliance for Coding Agents
- Cross-lingual Calibration of Pre-Generation Success Probes for Multilingual LLM Routing
- APOD: reasoning-guided agentic population ordinary differential equation discovery for pharmacological digital twins
- SO(3)-RoPE for Spherical Transformers
- Joint Class-Time Learning for Video Classification with Multi-Instance Partial-Label Learning
- Artificial Intelligence and the New Science of Culture
- Few-Shot Prototype Head Adaptation for On-Device ECG Personalization on PSoC~6
- CentriQ: Calibration-Free Quantization of Diffusion Transformers via Exact Mean Centering
- Ramp Metering Control via Hybrid State Deep Reinforcement Learning in Partially Observable Connected Vehicle Environments
- VLA-ZO: Fast Zeroth-Order Adaptation for Vision-Language-Action Models
- Future Anchored Verification and Online Recovery for World Action Models
- When Are Concept Bottleneck Model Explanations Faithful and Compact?
- GAMBIT: Learning to Plan Continuous Multi-Robot Trajectories
- DialectSentEval 2026: Arabic Dialect Sentiment Analysis and Swapping Shared Task
- Wiring Matters: Injection Topology and Initialization of Affordance Heads in Vision-Language-Action Policies
- Agentic schema-guided extraction of materials process knowledge from scientific literature
- Fine-Tuning a 3B-Parameter LLM on a Smartphone: Characterizing Sustained Training
- MeSD: Multi-Evidence Self-Distillation for VideoLLM
- Steering by Influence: Curvature Aware Data Weighting for Activation Steering
- Scaling Down the Scaling Laws: Parameter Efficiency and Compute-Optimal Training in Resource-Constrained Large Language Models
- RAISED: Self-Distillation for Robustness to Prompt Injection in LLM Agents
- Quantifying the Stability of Multi-Step Reasoning via Error Amplification
- SpatialChain: A Benchmark for Auditing Spatial Reasoning Faithfulness in VLMs
- Toward Reliable Infant Pose Estimation: A Training-Dynamics Approach to Noisy Annotation Detection
- SoK: Semantic Decision Engines in Network Control Loops
- Valid Stopping in Adaptive Generator-Verifier Loops
- Better Call Reward: Reward Hacking as Strategic Abstention in Legal Reasoning Models
- ARO: Aligned Representation learning for multi-Omics data
- Choosing an energy-efficient software architecture for building system diagnostic support
- The Assistance Dilemma: Learning to Teach via Multi-Turn Reinforcement Learning
- A Physics-Guided Transformer Framework for Electromigration Analysis in Multi-Segment Interconnects
- Latent Flow Matching for Molecular Graph Generation
- Odyssey: A Closed-Loop Benchmark for Long-Horizon Real-World Driving with Explicit Navigation Routes
- Topology-Informed Prompt-Conditioned Universal Segmentation of Uterine Structures from Ultrasound and MRI
- Harmful Content Generation in Text-to-Image Models: Capabilities and Moderation Limitations
- ArtifactArena: Evaluating Models by What They Build in the Physical World
- GCTAuto-encoder: A Cross modal Framework for Security Flaw Detection in IoT Networks
- Anatomy of LLM Sycophancy: What a Flip Rate Hides
- Symmetry and AI-assisted discovery of magic-state factories
- BrainTRACE: Tracing Longitudinal, Multimodal, and Volumetric Evidence in Brain MRI Clinical Reasoning
- Does AI Help Cyber Attackers or Defenders? Evidence from Nonpublic Vulnerabilities and Subsequent Attacks
- Mind the Accent Gap: British Accent Robustness in Speech-Driven Financial Voice Assistants
- SimForcing: Distilling Simulation Motion Priors into Real-Domain Robot World Models
- Word-Level Text Unmixing via Evidence-Preserving Ownership Routing with Language Models
- Frozen Factor or Spectral Band? Disentangling Two Choices in Low-Rank LoRA
- Large Language Model-Guided Discovery of Weight-Five Bivariate Bicycle Codes
- Differentially Private Mixing of Public Datasets Improves Private Learning
- AffordCraft: Scalable Construction of Task-Ready Simulation Assets from Single Images
- VideoTapestry: Query-Adaptive Memory Refinement for Multi-Agent Long-Video Understanding
- Closing the Context Gap: Activation Alignment for Tabular In-Context Learning
- BRANCH-MoE: Balance-Aware Tree Routing for Large Embedding Models
- BazaarBench: Delegation Safety in Decentralized C2C Marketplaces Run by LLM Agents
- Balancing Memory Pathways: Analyzing and Improving Memory Utilization in Hybrid LMs
- MatrixFormer: A Foundation Model for Matrix Completion
- IdeaLens: Detecting AI Ideas in Long-form Writing
- MC-Sparse: Deconstructing and Closing the Dense-Sparse Attention Gap in Diffusion Transformers
- Sharpen Without Search: On-Policy Distillation of Sequence-Level Power Distribution
- TAPDreamer: Transferable Adversarial Patches for World Action Models
- Paradee: Distilling Kokoro-82M into an 8M-Parameter Single-Voice Text-to-Speech Model
- Deep Learning for Sleep Heart Rate Estimation from Accelerometers: Toward Population-Scale Cardiac Insight Without Optical Sensors
- CLIFT: Conformal Self-Verification for Web Agent Training and Test-Time Scaling
- MemPilot: Orchestrating On-Demand Multimodal Memory Curation for LLM Agents
- UniSlider: Perceptually Uniform Sliders for Continuous Image Editing
- Recursive Video In-Context Learning for Agentic Robot
- Learning to Read the Contextual Tokens in Diffusion Transformers
- Base Models Can Reason By Taking a Cue From Training Data
- One Figure, Every Canvas: Editable Flowchart Relayout via Agentic Pipeline
- GFPack++: Attention-Driven Gradient Fields for Optimizing 2D Irregular Packing
- Answer Set Networks: Casting Answer Set Programming into Deep Learning
- Towards LLM Agents for Earth Observation
- OpenPhone: Mobile Agentic Foundation Models
- PaTAS: A Framework for Trust Propagation in Neural Networks Using Subjective Logic
- An Empirical Study of On-Device Translation for Real-Time Live-Stream Chat on Mobile Devices
- Neutral Substrates: A Design Constraint for Shared Records Under Persistent Interpretive Disagreement
- When Is Enough Not Enough? Illusory Completion in Search Agents
- AI Mental Models: Learned Intuition and Deliberation in a Bounded Neural Architecture
- ReLope: From Hidden-State Probing to a Decision Module for Multimodal LLM Routing
- The Ultimate Tutorial for AI-driven Scale Development in Generative Psychometrics: Releasing AIGENIE from its Bottle
- AI Assistance Reduces Persistence and Hurts Independent Performance
- High-Precision Estimation of the State-Space Complexity of Shogi via the Monte Carlo Method
- DERM-3R: A Resource-Efficient Multimodal Agents Framework for Dermatologic Diagnosis and Treatment in Real-World Clinical Settings
- Retrieval-Augmented Generation Must Move Beyond Factual Grounding to Represent Diverse Opinions
- From Tool to Agent: How Worker Needs Reorganize Across the Agentic Roles of Workplace AI
- EnvSimBench: A Benchmark for Evaluating and Improving LLM-Based Environment Simulation
- Agentic AI Scientists Are Not Built For Autonomous Scientific Discovery
- PiCA: Pivot-Based Credit Assignment for Search Agentic Reinforcement Learning
- MAGE: Multi-Agent Self-Evolution with Co-Evolutionary Knowledge Graphs
- An Agentic AI Framework with Large Language Models and Chain-of-Thought for UAV-Assisted Logistics Scheduling with Mobile Edge Computing
- Inference-Time Amplification of Weak Reasoning Models
- Grounded Continuation: A Linear-Time Runtime Verifier for LLM Conversations
- From Table to Cell: Attention for Better Reasoning with TABALIGN
- Nudging Beyond the Comfort Zone: Efficient Strategy-Guided Exploration for RLVR
- ECG-WM: A Physiology-Informed ECG World Model for Clinical Intervention Simulation
- Agentic Trading: When LLM Agents Meet Financial Markets
- Towards Direct Evaluation of Harness Optimizers via Priority Ranking
- Dr-CiK: A Benchmark testing Deep Research for Context-Aided Forecasting
- Plant, Persist, Trigger: Sleeper Attack on Large Language Model Agents
- MEMENTO: Leveraging Web as a Learning Signal for Low-Data Domains
- Characterizing How Complex Agentic AI Systems Handle General Tasks: A Trace-Based Simulation Study
- Beyond Objective Equivalence: Constraint Injection for LLM-Based Optimization Modeling on Vehicle Routing Problems
- Scaling Participation in Modular AI Systems
- Safety is Contextual, LLM-Judges Are Not: Navigating the Rigid Priors of Evaluators
- CIAware-Bench: Benchmarking Control Intervention Awareness Across Frontier LLMs
- SelfEvoSkill: Paired Execution Audits for Skill Revision without External Outcome Supervision
- RTSGameBench: An RTS Benchmark for Strategic Reasoning by Vision-Language Models
- Agent MechSuits: Mechanistic Subspace Safety Steering for Multi-Turn CLI Agents
- DART: Draft-Agreement Routing for Training-Free Adaptive Thinking Budgets in Hybrid Reasoning Models
- TIPS: Topological Ill-Posedness Probing and Steering in Large Language Models
- MORPH: Generative Retrieval via Diffusion Transformer with Metric-Ordered Sequence Training and Hybrid-Policy Preference Optimization
- Look-Before-Move: Narrative-Grounded World Visual Attention in Dynamic 3D Story Worlds
- Beyond Compilation: Evaluating Faithful Natural-Language-to-Lean Statement Formalization
- Mastermind: Strategy-grounded Learning for Repository-Scale Vulnerability Reproduction
- Subliminal Clocks: Latent Time Modelling in Diffusion Language Models
- AgenticPD: A Stage-Aware Agentic Framework for Closed-Loop Physical Design Optimization
- The Harness Effect: How Orchestration Design Sets the Token Economics of Enterprise Agentic AI
- Measuring Intelligence Beyond Human Scale
- Rethinking the Evaluation of Harness Evolution for Agents
- What Keeps Vision-Language Models Looking at the Image?
- Cura 1T: Healthcare Foundation Model via Recursive Self-Improvement
- A Dual-Hypothesis Reasoning Framework for LLM Guardrails
- Neuro-Symbolic Meta-Policies for Temporal Knowledge-Graph Memory under Partial Observability
- TILT: Model-Intrinsic Reward Alignment For Compositional Diffusion
- Deconstructing Off-Policy Ratios: Entropy-Normalized Trust Regions for Asynchronous Reinforcement Learning
- Towards Trustworthy Physical Intelligence: From Theory to Practice Across Life Cycle
- Dimensionality and Measurement Precision in HLE's Multiple-Choice Subset
- SKILL-KD: Contrastive Skill Distillation for LLM Agents
- RA-CAD: Learning Post-Execution Critique for State-Aware Text-to-CAD Generation
- NxN E-valuation: Hypothesis Certification via a Conformal CRT Null
- The Value of a Prompt: An LLM-Relative Kolmogorov-Complexity Approach
- CTIFoundry: An Agent-Native Corpus Scaffold for Cyber Threat Intelligence
- Hidden in the Request: Explaining Unethical LLM Compliance through Token Relevance
- Towards Reliable, Generalizable, and Specific In-Context Knowledge Editing via Multi-Objective Reinforcement Learning
- RealSWE: A Compositional Evaluation of Coding Agents under Realistic User Requests
- SEPO: Evidence-Grounded Prompt Optimization via Structural Editing
- DramaChain Bench: An End-to-End Benchmark for Short-Drama Generation
- Can LLMs Discover Scientific Laws in Real and Parallel Worlds?
- Answer-Distribution Trajectories: A Stochastic-Dynamics View of LLM Reasoning
- A Function-Space Approach to the Statistical Mechanics of Learning Dynamics
- Reading the Whole Heart: Latent-Attention Masked Autoencoders for Multimodal Cardiac Representation Learning
- Schizophrenia Detection from EEG Signals: A Transformer Framework with Spectrogram Representation
- EnterpriseVal: Quantifying the Efficacy, Reliability and Value of Generative AI in the Enterprise
- When No One Owns the Judgment: Accountability Under Contribution Dissolution in Human-AI Collaboration
- SWE-Prometheus: Measuring Engineering Governance Improvements in Real-World Repositories
- AtomWorld-Mem: Memory-Restored World States for Long-Horizon Atomistic Evolution
- Dude, Where's My State? Execution Information Requirements for Stateful Agents
- FedMHAR: Federated Multimodal Human Activity Recognition using Multi-Agent Reinforcement Learning
- Robust Biomolecular Complex Design Across Protein Conformational Landscapes
- Curating Merchant-Matching Training Data with Two Confidence-Gated Local LLM Judges
- Jev in Medicine: A Benchmark Evaluation
- ReplayLens: Auditing Agents' Use of Outcomes
- PhoneCLI: From App Interfaces to Callable Commands for Mobile Agents
- Learn from the Gap: Differential-Aware Advantage Pruning with Adaptive Rollout Sampling for GRPO
- ReMem: Rethinking Perception and Memory in Long-Context Recommendation Agents
- EnterpriseBench: Benchmarking LLM Agents on Enterprise-Level Strategic Reasoning and Decision-Making
- Adam under Generalized Smoothness with Second-Moment-Type Stochastic Gradients
- Advancing Entropy-Level Credit Assignment in RLVR via Proximal Entropy Policy Optimization
- ChronoGraph: Functional 4D Scene Graphs with Vision-Language Models for Interaction Understanding and Grounded Planning
- Safety of Latent Communication in Multi-Agent Systems
- Pay for the Fault, Not the Flow: Label-Free In-Flow Multi-Agent Workflow Optimization
- PRISM: A Category-Theoretic Framework for Measuring and Refining Multimodal Analogies
- Rethinking Probability-Based Reinforcement Learning From Posterior Concentration
- RealCompanion: Benchmarking Human Understanding from Reasoning over Longitudinal Real-World Conversations
- Mitigating Social Sycophancy via Pluralistic Preference Optimization
- Geometry Meets Physics: Data-Efficient Pre-Training for Unstructured Neural PDE Solvers
- Multilingual GSM-Symbolic: What determines capability transfer across languages?
- Understanding Certified Training with Interval Bound Propagation
- PPFedIT: Towards Privacy-Preserving Federated Instruction Tuning with Few-shot Local Examples
- CHAOSMINING: Benchmarking Post-Hoc Attribution with Sparse Informative Features in High Dimensions
- Efficient Hierarchical Transformers for Representing Log Data
- HardCore Generation: Generating Hard UNSAT Problems for Data Augmentation
- Screw Geometry Meets Bandits: Incremental Acquisition of Demonstrations to Generate Manipulation Plans
- MoFlow: One-Step Flow Matching for Human Trajectory Forecasting via Implicit Maximum Likelihood Estimation based Distillation
- 1000 Layer Networks for Self-Supervised RL: Scaling Depth Can Enable New Goal-Reaching Capabilities
- Optimizing Breast Cancer Detection in Mammograms: A Comprehensive Study of Transfer Learning, Resolution Reduction, and Multi-View Classification
- Robotic Long-Horizon Manipulation with Progressive In-Context Code Generation and Episodic Feedback
- Robotic Long-Horizon Manipulation with Bayesian Non-parametric Skill Priors
- Erased but Not Forgotten: How Backdoors Compromise Concept Erasure
- A Survey of Robotic Navigation and Manipulation with Physics Simulators in the Era of Embodied AI
- Graph-Based Floor Separation Using Node Embeddings and Clustering of WiFi Trajectories
- Using Echo-State Networks to Reproduce Rare Events in Chaotic Systems
- Anomaly Detection and Generation with Diffusion Models: A Survey
- EduAlign: Aligning Educational LLMs for Helpfulness, Personalization, and Creativity
- OCSVM-Guided Representation Learning for Unsupervised Anomaly Detection
- Resource-Limited Joint Multimodal Sentiment Reasoning and Classification via Chain-of-Thought Enhancement and Distillation
- Enhancing Knowledge Tracing through Leakage-Free and Recency-Aware Embeddings
- ZeST: an VLM-based Zero-Shot Traversability Navigation for Unknown Environments
- COCORELI: Enforcing Execution Preconditions for Reliable Collaborative Instruction Following
- Physics-informed GNN for medium-high voltage AC power flow with edge-aware attention and line search correction operator
- PATCH: Learnable Tile-level Hybrid Sparsity for LLMs
- Reinforcement Learning with Verifiable yet Noisy Rewards under Imperfect Verifiers
- Neural Bayesian Filtering
- Inverse Mixed-Integer Programming: Learning Constraints then Objective Functions
- FreshBrew: A Benchmark for Evaluating AI Agents on Java Code Migration
- Demystifying Hybrid Thinking: Can LLMs Truly Switch Between Think and No-Think?
- MCA: Modality Composition Awareness for Robust Composed Multimodal Retrieval
- A Model Can Help Itself: Reward-Free Self-Training for LLM Reasoning
- Learning Granger Causality under Latent Confounding via Intervention-Induced Heterogeneity
- User Misconceptions of LLM-Based Conversational Programming Assistants
- MedForj: An open, large-scale foundational generative prior for high-resolution 3D brain MRI
- Oolong: Evaluating Long Context Reasoning and Aggregation Capabilities
- DeNoise: Learning Robust Graph Representations for Unsupervised Graph-Level Anomaly Detection
- ProDER: A Continual Learning Approach for Fault Classification and Localization in Evolving Smart Grids
- Scalable Decision Making for Games of Imperfect Information
- Global Optimization on Graph-Structured Data via Gaussian Processes with Spectral Representations
- Bi-objective chance-constrained evolutionary optimization for large-scale open-pit mine scheduling under geological uncertainty
- EulerESG: Automating ESG Disclosure Analysis with LLMs
- Hierarchical Decentralized Multi-Agent Coordination with Privacy-Preserving Knowledge Sharing: Extending AgentNet for Scalable Autonomous Systems
- Mixed Data Clustering Survey and Challenges
- The Universal Weight Subspace Hypothesis
- Towards Optimal Valve Prescription for Transcatheter Aortic Valve Replacement (TAVR) Surgery: A Machine Learning Approach
- Contrastive Time Series Forecasting with Anomalies
- Memory as Resonance: A Biomimetic Architecture for Infinite Context Memory on Ergodic Phonetic Manifolds
- Symphonym: Universal Phonetic Embeddings for Cross-Script Toponym Matching
- ATOD: An Evaluation Framework and Benchmark for Agentic Task-Oriented Dialogue Systems
- Vulnerability of LLMs' Stated Beliefs? LLMs Belief Resistance Check Through Strategic Persuasive Conversation Interventions
- FlashMoE: Reducing SSD I/O Bottlenecks via ML-Based Cache Replacement for Mixture-of-Experts Inference on Edge Devices
- De-occluding broadband metalens
- Routing-Aware Safety Alignment for Mixture-of-Experts Models
- VIRENA: Virtual Arena for Research, Education, and Democratic Innovation
- Same Pieces, Different Servers: A Tetris Benchmark for AI Agents as Served
- Dissecting Quantization Error: A Concentration-Alignment Perspective
- RAC: Rectified Flow Auto Coder
- Towards Robust Speech Deepfake Detection via Human-Inspired Reasoning
- COMIC: Agentic Sketch Comedy Generation
- You've Got a Golden Ticket: Improving Generative Robot Policies With A Single Noise Vector
- Multi-Agent Reinforcement Learning Counteracts Delayed CSI in Multi-Satellite Systems
- Beyond TVLA: Anderson-Darling Leakage Assessment for Neural Network Side-Channel Leakage Detection
- DeepStock: Reinforcement Learning with Policy Regularizations for Inventory Management
- WebNavigator: Global Web Navigation via Interaction Graph Retrieval
- MDKeyChunker: What Does One LLM Call per Chunk Buy for Markdown Retrieval?
- Pre-Deployment Complexity Estimation for Federated Perception Systems
- Sven: Singular Value Descent as a Computationally Efficient Natural Gradient Method
- An AI Teaching Assistant for Motion Picture Engineering
- $\pi^2$: Structure-Originated Reasoning Data Improves Long-Context Reasoning Ability of Large Language Models
- A Self-Calibrating Framework for Analog Circuit Sizing Using LLM-Derived Analytical Equations
- On the use of evolutionary optimization for the dynamic chance constrained open-pit mine scheduling problem
- Evaluating a Layered Prompt-Injection Defence for the Model Context Protocol: A Record-Level Audit of Decision Conventions, Corpus Provenance and Reproducibility
- PSI-Bench: Interpretable and Clinically Meaningful Evaluation of Depression Patient Simulators
- VIDA: A Dataset for Visually Dependent Ambiguity in Multimodal Machine Translation
- Learning to Theorize the World from Observation
- Active Learning for Communication Structure Optimization in LLM-Based Multi-Agent Systems
- Is Escalation Worth It? On the Depth of LLM Cascades
- PLOT: Progressive Localization via Optimal Transport in Neural Causal Abstraction
- Mask2Cause: Temporal Causal Discovery Beyond Causality in Mean
- Your Language Model is Its Own Critic: Reinforcement Learning with Value Estimation from Actor's Internal States
- The Endogeneity of Miscalibration: Impossibility and Escape in Scored Reporting
- TIDES: Implicit Time-Awareness in Selective State Space Models
- NCO: A Versatile Plug-in for Handling Negative Constraints in Decoding
- When Can Digital Personas Reliably Approximate Human Survey Findings?
- Premover: Fast Vision-Language-Action Control via Early Execution During Instruction Delivery
- Large Language Models for Agentic NetOps and AIOps: Architectures, Evaluation, and Safety
- AgentLens: Revealing The Lucky Pass Problem in SWE-Agent Evaluation
- Analog RF Computing: A New Paradigm for Energy-Efficient Edge AI Over MU-MIMO Systems
- Text Knows What, Tables Know When: Clinical Timeline Reconstruction via Retrieval-Augmented Multimodal Alignment
- GraViti: Graph-Level Variational Autoencoders with Relaxed Permutation Invariance
- An Assessment of Human vs. Model Uncertainty in Soft-Label Learning and Calibration
- PopPy: Opportunistically Exploiting Parallelism in Python Agentic Workflows
- Less Back-and-Forth: A Comparative Study of Structured Prompting
- TextReg: Mitigating Prompt Distributional Overfitting via Regularized Text-Space Optimization
- Transcoders Trace Visual Grounding and Hallucinations in Vision-Language Models
- The TIME Machine: On The Power of Motion for Efficient Perception
- DrawVideo: Grounded and Faithful Multi-Shot Video Generation from Storyboard Keyframe Sketches
- EchoDistill: Robust Large Audio Language Models via Noisy-to-Clean Self-Distillation
- Measuring the Depth of LLM Unlearning via Activation Patching
- A Controlled Synthetic Benchmark for Educational Aspect-Based Sentiment Analysis
- Erased but Exploitable: Black-box Embedding-Aware Prompting Against Unlearned Text-to-Image Diffusion Models
- vSV-ViT: Variable-size SuperVertex Vision Transformer for Cortical Surface Learning in Alzheimer's Disease
- CiteCheck: Retrieval-Grounded Detection of LLM Citation Hallucinations in Scientific Text
- Brain-IT-VQA: From Brain Signals to Answers
- EUDAIMONIA: Evaluating Undesirable Dynamics in AI
- SWIM: Compact Environment Representation for Reinforcement Learning Human Swimming
- GEM: Geometric Erasure by Contrastive Velocity Matching in Rectified Flows
- EPIC: Efficient and Parallel Inference under CFG Constraints for Diffusion Language Models
- Low-Resource Safety Failures Are Action Failures, Not Representation Failures
- All you need to break LLMs are Black-Box, Adapting, Efficient, Transferable, Harmful, Applicable ... Attacks
- A Goal-Set Characterization of Task Composition in the Boolean Task Algebra
- Benchmarks in Leipzig
- LatentWave: JEPA Pretraining for Wireless Foundation Models
- MMBU: A Massive Multi-modal Biomedical Understanding Benchmark to Probe the Perception Capabilities of Vision-Language Models
- STREAM: Stochastic Riemannian Flow Matching with Anisotropic Decoder for Digital Histopathology Image Generation
- CultureScore: Evaluating Cultural Faithfulness in Video Generation Models
- LEAF: Growing Trees Without Branching for Speech-Aware Large Language Model Post-Training
- Decentralized Multi-Agent Systems with Shared Context
- Gefen: Optimized Stochastic Optimizer
- Machine learning modeling of hit-into-play probability and quantification of pitch sequence effects
- EG-VQA: Benchmarking Verifiable Video Question Answering with Grounded Temporal Evidence
- Narrative-UFET: Narrative Generation for Ultra-Fine Entity Typing
- Behavior Uncloning: Distilling Mode Redirection into Policy Weights without Inference-Time Steering
- Multi-Level Distributional Entropy from Flow Summary Statistics for Explainable Network Intrusion Detection
- Anchored or Drifting: What Recursive Self-Generation Reveals About Training Data
- EgoSafetyBench: A Diagnostic Egocentric Video Benchmark for Evaluating Embodied VLMs as Runtime Safety Guards
- Meta-Transfer Learning for mmWave Beam Alignment
- From Personas to Plot: Character-Grounded Multi-Agent Story Generation for Long-Form Narratives
- Office Comprehension Benchmark
- AnyGroundBench: A Multi-Domain Adaptation Benchmark for Video Grounding in VLMs
- Safe Inference-Time Alignment via Lagrangian Reward Augmentation
- Mask2Real-WM: Controllable Dexterous World Models via Segmentation Masks as a Sim-to-Real Bridge
- UI2App: Benchmarking Visual Interaction Inference in Executable Web Application Generation
- Co-LMLM: Continuous-Query Limited Memory Language Models
- Authority-Bound Governance of Heterogeneous AI Security Decisions in Telecom and IoT Networks
- Exact and Certified Data Shapley for Weighted Nearest Neighbor Regression
- PFAdapter: Hierarchical LoRA Decomposition for Personalized Federated MLLMs
- Continual Learning with Elastic Regularization and Synthetic Replay for Federated MLLM Fine-Tuning
- Memory as a Controlled Process: Learned Adaptive Memory Management for LLM Agents
- CAS I: A Geometric Coding Theorem
- SLPO: Scaling Latent Reasoning via a Surrogate Policy
- Medical foundation models converge less under label supervision
- Looping Is Not Reliability: State-Bound Evidence and Typed Revision Contracts for Agentic Code Repair
- When Synthetic Users Fail: A Cross-Domain Benchmark of LLM-Simulated Human Survey Responses
- The Implementation Lottery: Auditing Idea Reliability in Automated Research
- Arm2Air: Cross-Embodiment Skeleton Transfer for 3D Relay Formation
- Cost-Effective Automated Judging of Natural-Language Mathematical Proofs
- GraRe: Grasp Candidate Re-Ranking for Frozen 6-DoF Grasp Detectors
- Training nGPT
- MemArena: An Ego-Centric Benchmark for On-Device Agentic Personal Memory Assistants at Scale
- CLEARMIND: Closed-Form Laplacian Embeddings of Arakelov-Green Resistance for Morphological Intrinsic Neuronal Distances
- Site Is Decodable Before Pretraining: Negative Controls for Probing Frozen Brain-MRI Foundation Models
- The AI Accountability Ecosystem in the Era of Language Models
- Learn What's Left, Not What's Mastered: Saturation Aware Advantage Reweighting for Multi-Reward Policy Optimization
- The Acknowledgment Point Is the System: Durable Policy-Decision Receipts for AI Audit Evidence
- Science Done on a Machine by a Machine: AI Agents in Computational Chemistry
- TRACE: Training-time Report-guided and Clinically Ordered Concept Editing
- Does Fine-Tuning Undo Activation Steering? Behavioural Recovery Without Weight-Edit Reversal
- Pointing the Way, Hiding the Destination: Practical Private Dense Retrieval at Scale
- Difference-in-Differences on a Censored Rating Scale Can Manufacture an Effect: Evidence from a Pre-Registered LLM-Judge Audit
- The Hallucination Signal Is a Mean Shift: Why Simple Probes Suffice
- CPR for LLMs: Critical-Point Routing against Catastrophic Forgetting in Domain Adaptation
- SIR: Self-improving Red-teaming for Compute Use Agents
- Capability-Gated Language Models: Security Composes, Utility Does Not
- From Fixed Keys to Readable Schemas: Small Language Models for Vehicle Agent Function Calls
- A Measured Communication-Quality Frontier for Federated LoRA Fine-Tuning with Adaptive Phase-Switching
- Data-Free On-Policy Distillation: How Far Can We Go Without External Data?
- TriCalRAG: A Three-Strategy, Retrieval-Augmented Benchmark for On-Premise LLM-Based Root Cause Analysis in AIOps
- CSWAM: Better Causal Semantic Representations for Out-of-Distribution Generalization in World Action Models
- Beyond Outcomes: Dual-View Relational Learning for Efficient Agent Benchmarking
- Semantic Action Graph: A Shared Representation for Agent Grounding and Human Interpretation of Sports Highlights
- VidOmni-Bench: A Benchmark for Fine-Grained Video Understanding via Spatio-Temporal Event Verification across Complexity and Duration
- The Law of Stop: Interruptibility, Injunctions, and the Governance of Agentic AI
- RRSI: Regularized Recursive Self-Improvement of Agent Harnesses
- TraceVIC: Causal Reasoning over Code Evolution for Identifying Vulnerability-Inducing Commits
- Technical Manual for Toolkit for Confidence-Corpus Consistency, Corpus Absorption and Rule Learning via Fine-Tuning on a Fabricated Corpus
- Rolling-WAM: World Action Models with Rolling Imagination
- A Unified Account of Concepts and Chunks
- FLIP: Final Layer Inference-Time Probing for Vision-Language Models
- The Linear Representation Hypothesis for Vision-Language-Action Models
- Activation Flow: Manufacturing Activations for Steering
- Timestep Weighting: A Hidden Key to Effective ELBO-Based Flow-Matching RL
- ASCEND: Personal AI Agents for Autonomous Scientific Computing Across HPC Clusters and GPU Workstations
- Evaluating System One Models for Agent Security Decisions: Reliability, Calibration, and Selective Automation
- Collaborative Synthetic Data for Privacy-Preserving Financial Fraud Detection Across Organizational Silos
- Counterfactual Rollout Replay: Forkable Environments as Free Process Rewards for Software Engineering Agents
- Finite Probes Suffice: Identifiability and Universality for Weight-Space Learning
- LLMs are not stochastic parrots: Evidence for meaning-mediated abstraction from conlang-like tasks
- Calibrated to Whom? Persona and Language Effects on Cultural Values in JEV
- Distilling Diffusion Score Discrepancy for Efficient Training Data Attribution
- False Frontiers: Diagnosing and Mitigating Co-Cheating in Self-Evolving Search Agents
- UniAE-MoE: A Unified Audio Encoder via Mixture of Experts
- Wavelet Flow Matching for Time Series
- Free Everywhere, Exact on Trees: PPO's Dropped Correction Buys Sample Efficiency Under Aggressive Reuse
- CAS II: Orbits as Models: Kolmogorov's Structure Function under Symmetry
- Nous: Learning and Certifying Memory Decisions Before Source Calibration
- Four Ways to Grow a Classifier and Why One of Them Cannot Learn
- The Life Cycle of a Massive Activation: Stochastic Birth, Weight-Decay-Driven Growth, and Competitive Consolidation
- Memetic Trojans: Social Contagions as Carriers of Adversarial Payloads in Agent Networks
- PACT: End-to-End Learning of Human Pose, Contacts, and Forces from Video
- SHARPO: Segment-Level Credit Assignment for Agentic Reinforcement Learning
- RelationVGGT: Visual Geometry Transformers for 3D Spatial Relation Segmentation
- CineMR: Tool-Integrated Vision-Language Reasoning for Quantitative Cardiac MRI Assessment
- Removing spurious minima for planar features by skip connections
- Generalization Bounds for Flow-matching Generative Models for Intrinsically Low-dimensional Data
- TEMPEST: Temporal Embeddings for Scalable Driver Identification via Angular Margin Learning
- When Does External Guidance Help LLM Reasoning? A Bias-Variance Theory of Guidance-Augmented GRPO
- Neutrosophic Ensemble Classification for Uncertainty-Aware Bearing Fault Detection: Evidence from Laboratory and Variable-Speed Industrial Benchmarks
- Comparative review of hybrid forecasting models for short-term prediction of building thermal load
- Learning When to Refine: Long-Horizon Reinforcement Learning for Budgeted Neural-Operator PDE Solvers
- Event-Driven ML Pipeline Orchestration for Manufacturing: An AWS Industry Experience
- Learning from Unreliable Trajectories: Adversarially-Robust Federated Q-Learning
- AttSVD:Prompt-Adaptive Low-Rank KV Cache Compression via Attention-Guided SVD
- Near-Optimal Sample Complexity for Recursive Entropic Risk Reinforcement Learning with a Generative Model
- QiYao-I: A Manifold Based Foundation Model for Irregular Multivariate Time Series Forecasting
- Learning to Remember: Distilling Memory Retention for Compact Recurrent Neural Networks
- Learning to Decide, Not to Reason: Parameter-Efficient Decision Operators via Low-Rank Activation Steering
- Do Neural PDE Solvers Learn the Right Dynamics?
- EVFormer: An Egocentric Vision-EMG Bidirectional Attention Model for Bimanual Hand Pose Estimation
- Uncertainty in Representation Learning on Knowledge Graphs
- A Data-Driven Framework for Unsupervised Monitoring of Transmission Systems Using End-of-Line Testing Data: A Case Study at Ford Motor Company
- An Information-Theoretic Evaluation Framework for Benchmark and Model Diagnosis in Knowledge Tracing
- Repair Lot Skyline: A Weighted Constraint Satisfaction Approach to Pavement Repair Optimization from Geospatial Hazard Density
- DART-ES: Difficulty-Aware Reweighting and Targeted Replay for Fine-Tuning LLMs with Evolution Strategies
- Mask-Guided KV Cache Eviction in Block Diffusion Language Models
- Should We Skip Diffusion?
- STOCK-JEPA: Prior-Anchored Latent Revision Representation Learning in Equity Markets
- Identifiable World Models from Pretrained Diffusion Representations
- Shaping the Wind: Nested Potentials for Kinematically Admissible Urban Wind Prediction
- The Premise Is the Problem: Exchangeability Failure in Self-Monitored Test-Time Adaptation
- TRIAGE: Direction-Aware Mismatch Stabilization of Native NVFP4 Reinforcement Learning
- Skillful Data-Driven Subseasonal Soil Moisture Forecasting: Prospects and Limits for Flash Drought Prediction
- Learning to Simulate Individuals from Macro Social Signals
- What Must Replay Preserve? Separating Correctable Bias from Class Correspondence
- Few-Shot Bioactivity Prediction with Meta-Learning under Assay Heterogeneity
- When Attention Does Not Explain the Peak: Temporal Reference vs. Forecast Output in Attention-Based Time-Series Forecasting
- SchemaFill: Efficient LLM Tool Calling via Slot-Parallel Speculative Decoding
- Muon Is Theoretically Wrong For Convolutions, But Empirically Effective
- LiLib: Lifelong Air-to-Ground Path-Loss Prediction on UAVs via a Drift-Triggered Model Library
- Sample-Optimal Estimation of the Fr\'echet Inception Distance
- Will the Judge Flip? Predicting Position-Sensitive LLM Judgments from Residual Stream Activations
- SoloQ: Calibration-Free Quantization for Diffusion Language Models
- The Implicit Bias of Hyperbolic Representation Learning for Multiclass Data: A Busemann Risk Perspective
- Adversarial Training for Deep Hedging in Nonstationary Markets
- A theory of platonic representations in language models
- CLM-as-a-Judge: Evaluating an Open Contrastive Decision Model on Public Judge Benchmarks
- Learning Scientific Exploration from Human Research Decision Trajectories
- Exact Unlearning via Quantized Sufficient Statistics
- Distributionally Robust Mixture-of-Experts Training
- Can LLM-assisted regularization increase forecast accuracy for migration flows in low data regimes?
- Reward-Driven Learning under Prompt-Level Differential Privacy
- Constant-Curvature Sliced Gromov-Wasserstein for Heterogeneous Cross-Curvature Alignment
- Data, Numbers, and Geometry: Three Tutorials on Numerical Methods, Machine Learning, and Evaluation
- Minimal Witness Reinforcement Learning
- Conditional Flow Matching for Transport Between Markov Processes
- Benchmarking Time Series Foundation Models for Load Forecasting Under Covariate Uncertainty
- Learning What to Distill: Bilevel Top-K Token Selection for Self-Distillation in Large Language Models
- Neural Fields Encode Adaptation Geometry
- Neural Algorithmic Reasoning for Graph Saddle Point Problems
- Algorithmically Aligned Neural Agglomerative Tree Construction
- Lock-in EP: An In-Situ Training Algorithm for Oscillatory Hardware
- FlexiFlow: Bandit-based Model Switching in ML Workflows
- ATLAS-AL: Adaptive Trust-Region for Latent Adversarial Searches via Active Learning
- Scale-Invariant Training for Time Series Foundation Models
- Structuring MoE Expert Selection for Agentic Reinforcement Learning
- Weight Oracles: Reading Neural Network Weights with Language Models
- Selective Critique for Cost-Aware LLM Agents in Long-Horizon Decision Making
- CausalBind: Causal Modeling and Learning for Protein-Molecule Virtual Screening
- Stepped MoE: Segment-Level Routing with Configurable Inference Complexity
- RELACE: retrospective likelihood-based action credit estimation for long-horizon language agents
- Towards Explainable Benchmarking for Data-driven Post-Wildfire Debris Flow Prediction
- Dynamic Budget Allocation for LLM Evaluation under Hard Resource Constraints
- Multigroup Fairness and Omniprediction: Separations and Equivalences
- Inference and learning in sparse autoencoders as natural gradient flow
- Fed-BRDECS: Privacy-Preserving and Heterogeneity-Aware Federated Deep Embedded Clustering
- What pass@k Cannot Measure: Evaluating Diversity and Capability Retention after Post-Training
- Evaluation of Active Feature Acquisition Policies with Tabular Foundation Models
- Learnable Spectral Activations
- Benchmarking Label-Revealed Online Updates for EEG BCI Decoding
- StaFIR: Convex Learning of Stationarity-Aware Causal Filters
- Decoupling What from Where: How Should a Small GUI Grounding Model Receive the Action Type?
- Fork-and-Flush: Escaping Idea Basins in Autoresearch Agents
- Active Feature Acquisition for Cost-Efficient Temporal Prediction with Reduced Participant Burden
- AlignQuant: Tile-Aligned Mixed-Precision Quantization for Efficient LLM Generation
- Interpretable Hypergraph Learning via Neural Additive Models
- Efficient Multimodal Inference through Adaptive Acquisition and Sequential Fusion
- Structure, Not Belief: Correlated Thompson Sampling from LLM-Derived Covariance in Combinatorial Semi-Bandits
- Adapting to Changes in Agent Behavior via Finite-Depth Policy Sensitivity
- SpecBraM: What Should an EEG Foundation Model Predict? Masked Band-Power Prediction versus Waveform Reconstruction
- Robust Importance Sampling for Rare Events via Constrained Gaussian Mixtures
- Who Bears the Burden? Learning Responsibility for Shared Constraints in Multi-Agent Reinforcement Learning
- Source-Learned Reliance for Selective Test-Time Adaptation of Multimodal Time Series
- Harmful SFT Leaves a Continuous Trace in LLM Checkpoint Updates
- Activation Denoising: A Robustness View on Parallel vs Sequential LLM Quantization
- Targeted search shows that random-device testing underestimates worst-case error in a simulated wave-based neural operator
- Preserving Unstable Modes Through Inverse Dynamics in JEPA World Models
- Foundation Model-Aided Multi-Agent Reinforcement Learning for Wireless Random Access Network Optimization
- Which and When to Admit: Gradient Admission for Data-Centric Small Language Model Finetuning
- Global Transport Couplings for Classifier-Free Guided Flows
- TAFFY: A Task-Adaptive Tabular Foundation Model with In-Context Diversity
- Learning a Mixture of GFlowNets
- Complementary Feature Domains: Information Preservation Does Not Imply Predictive-Contribution Preservation
- Mechanistic Interpretability of Atmospheric Rivers in GraphCast
- Hub for Outliers, Spokes for Inliers: Uniform Latent Space Construction for Dual-Mismatched Semi-Supervised Learning
- AFA-BANDIT: Provably Near-Optimal Online Multi-Feature Classification Under Budget Constraints
- Stateless Language Agents: Scaling Long-Horizon Automated Research
- Complementary Supervised and Self-Supervised Representations for Out-of-Distribution Graph Learning
- Does On-Policy Distillation for Safety Pose Backdoor Risks?
- MS-ECG-FM: Towards a More Universal Electrocardiogram Foundation Model for Health Monitoring using Multi-source Contrastive Learning
- Exact-Solution Volume and Length Generalization in Transformers
- Adaptive Model Inversion Attacks Generalize a Privacy-Robustness Tradeoff
- Improving Synthetic Data Generation for Argument Mining via Adversarial Reinforcement Learning
- WASD: Wasserstein-based Knowledge Distillation for Large Language Models
- Neuromotor Hierarchy Network: Physiological Inductive Biases for Robust Generalization in sEMG Decoding
- Cite What You Explore: Budget-Aware LLM Reasoning over Medical KGs with Verifiable Evidence
- Adversarially Trained Linear Transformers Are Optimal Robust In-Context Learners for Gaussian Mixtures
- TRACE: Rollout-Guided Quantization-Aware Training for FP4 Reinforcement Learning of MoE Language Models
- Towards One-for-All Foundation Model for Attributed Graph Clustering
- Extending Pathwise Gradients to Discrete Random Variables via Finite-Order Relaxation
- ServeLearnBench: How Well Can Agents Self-Improve from Serving Experience?
- The Geometry of Empowerment
- Adaptive Mean Estimation by In-Context Learning: A Gradient-Flow Analysis
- MASKerade: Token-Routed Mask Experts for Dense-to-MoE Upcycling
- SIFT: Search Intent-to-Filter Transformer for Multi-Task Personalized Filter Ranking at Airbnb
- $\alpha$Transfer: Coefficient Transfer for Efficient Model Merging
- TTNet: Multi-Task Deep Learning for Table Tennis Player Analysis with Smart Racket
- CANDLE: Cortical Null-Space Decomposition for Noninvasive Brain Source Imaging
- Retrieval Is Not Enough: Refreshing Memory for Frozen Time-Series Forecasters
- Privileged Context as Drift in On-Policy Self-Distillation
- Lost in the bf16 Cast: Exporting Ternary Language Models Can Revert Most Low-Learning-Rate Code Changes
- A Decision-Focused Neural Optimization Framework for Personalized Route Reproduction from Vehicle Trajectories
- Tram-FL: Reducing Communication and Computation Costs through Sequential Model Circulation in Decentralized Federated Learning
- On-Policy Distillation with Negative-Policy Rollouts
- Label-Efficient Deep Learning for ECG Delineation: A Multi-Dataset Benchmark against Widely Used Delineation Tools
- FC-SWE: Failure-Conditioned RL for Long-Horizon Software Engineering Agents
- Variance-Averse $n$-Step Offline Reinforcement Learning for Sparse Long-Horizon Environments
- ApexQuant: Data-Free Elastic Quantization by Residual Re-Isotropization
- Revisiting Temporal Regularization for Smooth Control in Deep Reinforcement Learning
- Generalized Matheron Variational Implicit Processes
- DecepEval: A Benchmark for Evaluating Deception in LLM Agents
- Learning a Ranking from Human Feedback in Log-Concave Random Utility Models
- Do Higher-Order Models Win for Higher-Order Reasons? Rethinking Performance Gains in Hypergraph Learning
- A Broader Look at Model Merging: Rethinking Implicit Regularization Induced by Task Arithmetic
- TICDA: Tabular In-Context Data Attribution
- Can phenotypic activity be predicted without experimental readouts?
- Spectra: Exact Component Transport for Test-Time Prior Adaptation in Simulation-Based Inference
- SepsisLens: Structure-Preserving Sequence Modelling for Decomposable Early Sepsis Warning
- A Riemannian Geometry for Low-rank Adaptation
- Detecting a Shift Is Not Enough: Exact Minimax Limits of Linear Representation Repair
- Optimization Encoders: Rethinking Second-Order Meta-Learning for Neural Fields
- Enhancing Diffusion Language Models with Autoregressive Post-Training Weights
- Attenuated in-context identification in time-series foundation models: diagnosis under counterfactual inputs and repair by synthetic forced-system fine-tuning
- Do LLMs Act on What They Know? From Partner Representations to Cooperative Actions
- Mu-DisCoCat: A Variational Pipeline for Compositional Generalization on Quantum Processors
- Symphony for Text Generation: Benchmarking Clinical Note Generation
- Align, Then Correct: Training-Free Two-Stage Low-Rank Compensation for Extremely Quantized Large Language Models
- On the Intrinsic Limited Robustness of Latent-Based Watermarking
- Finding the Heads and the Neurons Responsible for Network Information Retrieval in Language Models
- Reinforcement Learning with Segment Reward Feedback under Linear Function Approximation
- Performative Prediction with Selective Labels
- Scalable extraction and visualization of multi-attribute logical and functional dependencies in tabular data
- OxiGen: Oxidation-State-Aware Crystal Generation
- Structure-Aware Graph Abstention for Reliable Selective Forecasting
- Uncertainty Quantification Is Indispensable for Reliable Connectome-Based Graph Learning: A Narrative Review and Case Study
- Sensor Geometry as a Flow-Matching Prior for Multi-Channel Brain Signals
- Evolutionary One-Step Generators: Fast and Diverse Sampling for Discrete Design
- Accelerating the Development of PLGA In Situ Forming Depots Through AI-Driven Multi-Objective Optimization
- Decision-Focused Learning in MDPs: An Occupancy Measure Approach
- Atom-JEPA: Joint-Embedding Predictive Architecture for 3D Atomistic Systems
- VETTA: Coordinating Turn- and Token-Level Credit Assignment for Multi-Turn LLM Agents
- SSR: Sparse Segment Reduction for Ternary GEMM Acceleration
- Symmetry-Aware Feature Learning: A Polynomial Separation for Multi-Index Models
- Climbing the Design Ladder: Sequential Knowledge Distillation for Early-Stage Circuit Timing Prediction
- Learning PDE solution operators with variable initial conditions via Latent Dynamics Networks
- MetaLearnNCA: Few-Shot Offline Meta-Learning via Interacting Neural Cellular Automata
- PHBA: Prefix-State Hybrid Block Attention
- How Bregman Divergences Shape Shampoo
- FlowCF: Sparse Counterfactual Explanations for Mixed-Type Tabular Data using Flow Matching
- From Shared Demand Patterns to Local Uncertainty: Probabilistic Load Forecasting by Mixing Compact Adaptations
- DeltaTTT: Layerwise Optimization for Nonlinear Recurrent Memory
- Reinforcement Learning for Hierarchical Reasoning Rewards: Minimax-Optimal Rates with Transformers
- Valid for Free: Homophily-Gated Conformal Prediction for Training-Free Node Classification with Tabular Foundation Models
- Singular Value Decomposition: A Geometric Rediscovery, Where Proofs Become Algorithms
- Less Is More: A Leakage-Controlled Study of Dermoscopic Preprocessing for Joint Skin Lesion Classification and Segmentation with YOLO26
- How Learning Governs Unlearning across the Memorization-Generalization Spectrum
- Random Feature Gaussian Process Attention: Linear-Time Probabilistic Attention with Calibrated Uncertainty
- CNet: A Complex-Valued Deep Learning Framework with Wirtinger Autodifferentiation and FFT--Hadamard Convolution
- Multi-Label Perceptual Bug Detection in Video Games using Deep Learning on Gameplay Footage
- Early Memory Selection for Balanced Adam
- MemFLoRA: Memory-Floor LoRA for CNN Adaptation at the Edge
- Principled Under Pressure: Post-Training Decides Whether LLMs Act on Their Own Moral Judgment
- Variance-Optimal Off-Policy Evaluation with Conjunct Effect Modeling
- A Systematic Study of Small Language Models on Abstract Reasoning Tasks
- Probabilistic Counterfactual Inference for Discrete Outcomes in Gaussian-Process Causal Models
- GeneICL: A Tabular Foundation Model for Bulk Transcriptomics
- Optimal and Efficient Online Inverse Optimization
- On the Computational Tractability of Robust Bandits
- Reinforcement Learning with Conformal Action Sets: An Application to Sequential Recommendation
- Linear Bandits under Exact Sliding-Window Constraints
- Neural Petri flows for chemical reactions
- Conformal Prediction Sets Quantify Information Gain: A Theoretical Perspective
- The Geometry of Statistical Feature Learning in Mean-Field Langevin Dynamics
- Beyond Marginals: A Multi-Dimensional Evaluation Framework for Multi-Table Synthetic Data Generation
- Trajectools Demo: Towards No-Code Solutions for Movement Data Analytics
- Statistical Turbulence and High-Fidelity Disturbance Fields for Quadrotor Flight Control
- When Can World Models Recover Physical Laws?
- Zero-Shot Visualization: Exploring Text Corpora with User-Prompted Axes
- Axiom Satisfiability of Linear Rewards in Alignment
- Memory Prediction Excess: A Probabilistic Quantity for Predictive Gain and Memory Length in Stochastic Processes
- Component and Dimension Sparsity in Transformer Refusal Mechanisms
- Nonlocal Hamiltonian Dynamics on Sparse L\'evy Graphs: Spectral Analysis and Multimodal Sampling
- Anchor Divergence for Semantic Geometry in Contrastive Learning
- Extending Music Annotation Schemas: Zero-Shot Prediction or Few-Shot Adaptation?
- Low-Rank and Structured Sparse Tensor Decomposition for Anomaly Detection in Multivariate Functional Data
- FactorBench: A Portfolio-Aware Benchmark for Automated Factor Mining
- AdaLoop: Adaptive-Depth Latent Reasoning for Audio Language Models
- Principles that Guide, Actions that Inform: Agent Evolution via Knowledge Abstraction
- Hierarchy-GBP: Accelerating Factor Graph Inference via Abstraction and Recovery
- Where Does the Audio Jailbreak Live? A Controlled Frequency-Depth Audit of AdvWave-P on Qwen2-Audio
- Which Image Property Carries the Jailbreak? A Controlled Dissection of Image-to-Text Jailbreaks
- Calibrated Answers About Randomized Trials From a 4-Billion-Parameter Open Model: A Registered Test and a License-Clean Release
- Investigating Model Compression for Neural Machine Translation in the Biomedical Domain
- JIVEAdapter: A Multi-Task Additive Low-Rank Adapter via Joint and Individual Variation Explained (JIVE)
- Inference-Time Projection for Physically Valid Biomolecular Diffusion Models
- Data Fusion for Errors-in-Variables
- Behavioral Cloning Mystery
- sHAIL-Causal: A Sequential Staircase Procedure for Invariant Causal Predictor Discovery
- ImpactMat: Continuous Material Estimation for Inverse Impact Sound Rendering
- On Color Alignment in VAE Latent Spaces and Its Applications
- Learning Decision-Stump Thresholds in Context: Dynamics of Softmax Attention
- A Query Is Not a Commitment: Learning to Correct Expert Answers in Online Deferral
- Towards a Unified Misuse Monitoring Benchmark
- Evaluating Inference Compute for Generative AI: A Framework for Enterprise Workloads
- AMBER: Training Long-Horizon Web Agents through Append-Only Memory
- Jailbreaking Open-Weight LLMs via Random Embedding Perturbations
- Is this machine playing?
- CroissantMiner: Automated Extraction and Validation of Croissant Metadata for ML Datasets
- Interleaved Projected Gradient Descent for Safe Imitation Learning
- Learning Disentangled Representations with Quantum Variational Autoencoders
- An overview of machine learning-enhanced iterative methods for systems of linear and nonlinear equations
- How Inefficient Is Natural Gradient Descent? From Exact Optimality to \Theta ( \sqrt{ \log d } ) Divergence
- Hybrid Cross-Modal Attention Network for Early Breast Cancer Detection in Low-Resource Clinical Settings
- Lineage-Aware Memory Governance: A Derivation-Gated Framework for Privacy-Preserving Column-Level Access Control in Enterprise AI Agents
- What Words Keep of a Place: Zero-Shot Language Reasoning for Cross-View Geo-Localization
- A Single-Loop, Constant-Batch First-Order Penalty Method for Stochastic Bilevel Optimization
- Assumption-lean logistic regression with missing covariates
- Advantage of Entangled Learning Rules in Quantum Measurement Class Learning
- Logbook: Extremely Long-form Audio Event Understanding
- Evaluating Behavioral Context for Interpretable IAM Policy Risk Scoring in Cloud Environments
- Identity-Conditioned Score Fusion for Open-Set Person Re-Identification
- HyperNSDE: Personalized Neural SDEs for Joint Static-Longitudinal Clinical Data Generation
- DeepAJM: Deep Association Joint Model for Irregularly Sampled data
- A perspective note on likelihood approximation and inference for complex simulation models using a chain of aggregated normalizing flows
- Bayesian Optimization on Function Spaces via Sparse RKHS Manifolds
- AccentCL: Robust Accent Classification with Incremental Expansion
- Artifact removal improves electrodermal waveforms but not downstream classification in a virtual-reality balance task
- Deep Defence on Wheels: A Dual Intrusion Detection System Architecture for Comprehensive In-Vehicle Network Security
- Does Muon Need Fine-Grained Spectral Shaping?
- Two-Sample Testing for Random Graphs without Vertex Correspondence
- MobileVISTA: Generative Data Augmentation for Pose Generalization in Mobile Manipulation
- SkillFormer: Skill-Decomposed Adaptation for Audio Language Models
- Is $\sqrt{d}$ Separation Necessary for Gradient EM to Learn Gaussian Mixtures in High Dimensions?
- Seeing the Invisible: Physics-Guided Visual Prompting for Temperature- and Radiation-Aware VLA Navigation
- Two Vectors Replace In-Context Demos: Structured Task Adaptation via Embeddings
- CETUS: How Far Do Representations Trained on Earth Transfer to Cassini SAR of Titan?
- REViT-v2: Hierarchical Windowed Roto-reflection Equivariant ViT for Equivariant Feature Extraction
- Personal-Agent Mediated Recommendation with Cross-Platform User History
- Recurrent Looped Transformer
- BiGym 2.0: Benchmarking Learned and Agent-Developed Policies for Humanoid Household Manipulation
- The Robot Is Not Its Description: GaugeBench for Representation Robustness in Morphology-Aware Policies
- Modeling Latent Disturbances for Robust Decision-Making in World Models
- A Neural JKO Scheme for Hellinger-Kantorovich Gradient Flows via Monge-Growth Pairs
- Linear Fitness Subspace in Protein Language Models Enables Sample-Efficient Directed Evolution
- Learning Grasp Targeting from Point Clouds for Log Pile Clearing on a Hydraulic Crane
- Explore, Then Commit: Measurement-Efficient Scientific Law Discovery with Language Models
- Explicit Asymptotic Bounds for Sequential Calibration Beyond $T^{2/3}$
- Asymptotic Analysis of Empirical Risk Minimization on Entry-wise i.i.d. Heavy-Tailed Data
- Learning Explainable Representations of Complex Game-playing Strategies
- Towards the Automatic Synthesis of Interpretable Chess Tactics
- Uniform Discrete Diffusion Models are Minimax Optimal for Estimating Distributions with Small Effective Support Size
- Joint Workflow and Prompt Optimization for User Behavior Simulation
- CACHEFORGE: LLM-Guided End-to-End Generative Cache Replacement Policy for Performance and Hardware Efficiency
- Independent Multi-Agent Reinforcement Learning with Counterfactual Semantic-Social World Models
- Mathematical Invariant-Enabled Topological Neural Networks for Molecular and Materials Property Prediction
- Stability of Measure-to-Measure Transformers on Sub-Gaussian Data
- RefRoute: Decoupling Conditioning Cost from References via Compact Residual Conditioning and Spatial Routing
- Exact Calibration and Sharp Risk Geometry for Volume-Sampled Ridge Regression
- The Model Plants the Trigger: Answer-Side Backdoor Attacks in Multi-Turn Large Language Models
- SanSi: A Looped Typed Decision Model for System 1.5 Thinking
- Learning to Retrieve via Reinforcement Learning in Embedding Space
- Nash Social Welfare for Multi Armed Bandits: Trajectory-wise Expected and High Probability Regret
- High-dimensional online calibration from harmonic weights
- Trustworthy Method Comparison with AI Judges: Estimation and Design under Order, Batch, and Aggregation Effects
- APEX: Speculate smarter, not deeper
- Quantization Effects on Tool-Failure Recovery Vary Across Prompts and Evaluation Designs
- Persistent Memory in Multi-Agent LLM Inference: What It Costs, What It Buys, and When You Can Tell
- Common-Mode Errors Limit Low-Timestep Deep Spiking Q-Networks
- Stochastic Gradient Descent Ascent is Suboptimal for Nonconvex-PL Min-Max Games
- Do I Need the Cloud? Uncertainty-Aware Step-Level Handoff for Small Language Model Agents
- Forecast Accuracy Is Not Trading Profit: Evolving Small Recurrent Networks for Stock Return Prediction
- CHARTER: Auditing Reference Substitution in Hierarchical Compact-Evidence Evaluation for Computational Pathology
- Scen-Opt: A Scenario Optimization Toolbox for Data-Driven Convex Programming
- A self-learning scientific agent for X-ray diffraction
- ReFold: Training-Free Reversible Inter-Turn Context Folding for Long-Horizon Agents
- Learned Adaptive Multiresolution Diffusion Imaging
- Rethinking Faithfulness in LLMs: A Pairwise Context-Sensitive Perspective
- Isotropic Yet Undecodable: The Sequential Content-Sufficiency Gap in Latent-Predictive Text Representations
- Feature Encoding in VAE-based Audio Decoders: Effects of Input, Depth and Distribution
- FOSLS-deRhaNN: native de Rham neural classes for H(div) and H(curl) with applications to first-order system least-squares neural network methods for partial differential equations
- Learning consistent molecular mechanics force fields from first principles
- DAEDALUS: Bootstrapping Agent Memory from Self-Generated Tasks
- Language Carries the Expert's Impression: Instrument-Anchored LLM Judges Transfer Counseling-Quality Assessment and Beat In-Domain Training
- Self-Retrospection Distillation: Turning Post-hoc Experiences into Prior Foresight
- ProximalFM: Amortized Proximal Causal Inference under Hidden Confounding
- Explainable Rule Mining of IPv6 Extension-Header Presence Patterns from Paired-Vantage Captures
- Surviving the Router: Optimizing Skill Injections for Retrieval and Execution
- Energy-Aware Path Following: Comparative Analysis of Reinforcement Learning and NMPC for Electric Vehicles
- Beyond Waypoint Regression: Query-Based Cost Learning over Reachable Ego Futures for End-to-End Driving
- Beyond Marginal Monitoring: Distributed Joint-Distribution Testing for Data Concept Drift in Large Scale E-Commerce Operations
- Making COMET Comparable Across Scripts: Diagnosis and Correction of Tokeniser-Induced Script Bias in Indic MT Evaluation
- Beyond the Leaderboard: Multi-Dimensional Evaluation of Dense and Mixture-of-Experts Models for Automated Program Repair
- LFHE: Local-First Heuristic Evolution for Bounded Local Topology Search in Decentralized Learning with Non-IID Data
- Compact Robot Policies Need Fine-Grained Visual Representations
- Anytime-valid simulation-based hypothesis testing
- How Many Independent Samples Does a Satellite Image Contain? Generalization Bounds for Spatially Dependent Data
- LeanPlan: Optimal Planning with LLM-Generated Heuristics and Admissibility Proofs
- DySCo: Dynamic Sharding for Collaborative Edge-Cloud LLM Inference with Depth-Synchronized Batching
- Two-Sample Testing via Generative Processes
- Where Do Two Populations of Persistence Diagrams Differ? Calibrated Local Inference at a Fixed Budget
- CoDe-LoRA: Mitigating the Orthogonality Dilemma in Continual Learning of LLMs via Knowledge Consolidation and Decoupling
- The Standardization Trap: Certifying Joint Label Processing in Tabular Foundation Models
- Machine Learning for German Redispatch Forecasting under Data Delays and Temporal Distribution Shift
- DIPrune: Task-Aware Token Pruning with Dual Importance for Efficient Multimodal Language Models
- High-Dimensional Statistical Inference for Sparse Support Vector Machines
- Knowing When Not to Answer: Cross-Domain and Multi-Turn Generalization of Latent Underspecification Signals
- Agentic AutoRAG: RAG Pipeline Optimization through Reasoning-Driven Agents
- UNREAL: Unifying Retrieval and Long-Context with a Single Model
- Information-Dense Synthesis for Molecular Discovery
- X-OPM: Explainable Automatic Digital On-Chip Power Modeling for Enhanced Robustness
- Beyond Perturbation Magnitude: Direction-Dependent Responses in Multimodal Geometric Representations
- Toward Alignment Scaling Laws: A Framework and First Preregistered Measurements
- AnyBottle: A Recipe to Only Keep the Concepts You Really Need
- Systemization of Knowledge (SoK): Human-Centered AI Safety for Youth
- Latent space bias directions in LLMs capture confidence, not fairness
- Have I Seen Enough? Frozen Video-Language Models Encode Evidence Readiness
- RAG-PIBench: A Leakage-Aware Benchmark for Prompt-Injection Detection in Trustworthy RAG Systems
- FedDermaSeg: Federated Learning for Dermatological Image Segmentation
- Feature Information Dynamics in Diffusion
- SquidAgent: Parallelize Wisely, Coordinate Efficiently
- Steering Diffusion Models to Rare Events with Sequential Monte Carlo
- Secure Speculative Decoding for Large Language Models
- Prediction-powered inference for time series across space
- Co-Evolving Paths and Flows via Path-Flow Alignment
- When Forgetting is not Catastrophic: On the Mechanics of Spurious Forgetting
- Does an Agent's History Tell You When Compaction Will Hurt? A Modest, Bounded Effect on the TRACE Paired-Replay Corpus
- Denoising Hierarchical Representations: Joint Continuous Diffusion for Language Modeling
- Rapid Fredholm stabilization of the Kuramoto--Sivashinsky equation with unrestricted, spatially-varying anti-diffusion
- AdvSim2Real : Training Web Agents Against Adaptive Prompt Injection in a Web World Model
- QF3: Fast Flow RL with Filtered Q-Gradients
- Fast, Interpretable, and Deterministic Time Series Classification With a Bag-of-Receptive-Fields
- Probabilistic Truly Unordered Rule Sets
- FreDF: Learning to Forecast in the Frequency Domain
- Convergence of Sharpness-Aware Minimization Algorithms using Increasing Batch Size and Decaying Learning Rate
- Causal Effect Estimation under Networked Interference without Networked Unconfoundedness Assumption
- KO: Kinetics-inspired Neural Optimizer with PDE Simulation Approaches
- Time-o1: Time-Series Forecasting Needs Transformed Label Alignment
- Direct Regret Optimization in Bayesian Optimization
- Cross-Modality Controlled Molecule Generation with Diffusion Language Model
- PiERN: Token-Level Routing for Integrating High-Precision Computation and Reasoning
- HARL-A: An Extensible Benchmark Framework for Heterogeneous Multi-Agent Adversarial Reinforcement Learning in IsaacLab
- Predicting kernel regression learning curves from only raw data statistics
- DistDF: Time-Series Forecasting Needs Joint-Distribution Wasserstein Alignment
- Time-multiplexed layer reuse for physical neural networks
- Quadratic Direct Forecast for Training Multi-Step Time-Series Forecast Models
- Honesty over Accuracy: Trustworthy Language Models through Reinforced Hesitation
- Stabilizing Off-Policy Training for Long-Horizon LLM Agent via Turn-Level Importance Sampling and Clipping-Triggered Normalization
- Aligning LLMs with Biomedical Knowledge using Balanced Fine-Tuning
- A Footprint-Aware, High-Resolution Approach for Carbon Flux Prediction Across Diverse Ecosystems
- Optimal Transportation and Alignment Between Gaussian Measures
- Variational Physics-Informed Ansatz for Reconstructing Hidden Interaction Networks from Steady States
- Rethinking Fine-Tuning: Unlocking Hidden Capabilities in Vision-Language Models
- Stochastic Siamese MAE Pretraining for Longitudinal Medical Images
- Intersectional Fairness via Mixed-Integer Optimization
- Fast and Efficient Asynchronous Gossip Algorithm for Robust and Non-Smooth Convex Decentralized Learning
- Local exponential stability of mean-field Langevin descent-ascent and associated particle system
- Uncertainty Localization in LLM Reasoning via Embedding Perturbations
- RAM-Net: Linear-Time Sequence Modeling with Sparsely Addressable State
- UniST-Pred: A Robust Unified Framework for Spatio-Temporal Traffic Forecasting in Transportation Networks Under Disruptions
- Multi-Probe Zero Collision Hash (MPZCH): Mitigating Embedding Collisions and Enhancing Model Freshness in Large-Scale Recommenders
- Dual-Modality Multi-Stage Adversarial Safety Training: Robustifying Multimodal Web Agents Against Cross-Modal Attacks
- Federated Mixture-of-Experts Alignment on Mobile Edge Networks under Data Heterogeneity
- Process-Aware AI for Rainfall-Runoff Modeling: A Mass-Conserving Neural Framework with Hydrological Process Constraints
- Quality-Controlled Active Learning via Gaussian Processes for Robust Structure-Property Learning in Autonomous Microscopy
- Adaptive Semantic Communication for Wireless Image Transmission Leveraging Mixture-of-Experts Mechanism
- Neural Global Optimization via Iterative Refinement from Noisy Samples
- Sinkhorn doubly stochastic attention rank decay analysis
- Can an MLP Absorb Its Own Skip Connection Exactly?
- Observable Neural ODEs for Identifiable Causal Forecasting in Continuous Time
- From Dual Tracking to Clipping: Provably Faster Distributionally Robust Multi-Objective Optimization
- A Sharp Finite-Iteration Theory for Asynchronous Categorical Distributional Temporal-Difference Learning
- Particle Monte Carlo Tree Search
- MSPR: Multi-scale Predictive Representations for Goal-conditioned Reinforcement Learning
- Behavioral Guarantees for Proxy-Based Unlearning
- Stochastic Penalty-Barrier Method for Constrained Machine Learning
- To Call or Not to Call: Diagnosing Intrinsic Over-Calling Bias in LLM Agents
- Reinforcement Learning over Predictive Distributions for LLM Regression
- SeedER: Seed-Expand-Retrieve for Efficient Knowledge Graph Retrieval
- The Terminal Representation in Reinforcement Learning
- Task diversity produces systematic transfer but inhibits continual reinforcement learning
- Flow-Transformed Implicit Processes for Function-Space Variational Inference
- FFR: Forward-Forward Learning for Regression
- Mean-based algorithms: A lower bound and regret
- Express Language Modeling
- We Need Explanation Cards to Connect Explanation Algorithms to the Real World
- Explaining Attention with Program Synthesis
- Do Sparse Autoencoders Learn Meaningful Concept Hierarchies?
- Prime Fourier Embeddings: A Principled Basis for Modular Arithmetic
- Learning Probabilistic Filters with Strictly Proper Scoring Rules
- On the effectiveness of reward functions in reinforcement learning for confidence calibration of large language models
- Learning from Hindsight for VLA Reinforcement Learning
- Beyond Scaffold Splits: Structural-Frontier Evaluation Reveals Hidden Failures in ADMET Models
- Retrieval-Augmented Interpretable Learning: Towards Task-Specific Zero-Shot Models in Healthcare
- Variational-Ising-Attention:Tailored Attention Matters for Science
- Multi-Scale Structural Features for Continual, Comprehensible Visual Recognition in a Developmental Learning Framework
- EvoHarness-RL: Learning Runtime Harness Coordination for Self-Evolving Agents
- QUASAR: Lowering the Loss Floor of Quantization-Aware Training with Loss-Aware Reconstruction
- Reperesentation Geometry Matters for Planning with JEPA World Models
- PertMind: Eliciting Emergent Biological Reasoning in LLM via Reinforcement Learning on Cellular Perturbation Data
- Enhancing Distance-Based Graph Autoencoders with Structural Penalties for Dynamic Graph Embedding
- Systematic Evaluation of TabPFN-TS and Chronos-2 for Zero-Shot Heat Load Forecasting in District Heating Networks
- A Critical Audit of Spatiotemporal Forecasting Benchmark Datasets and Models
- Mapping the Emergence of Regularization-Driven Dynamics in Grokking
- The Latent Diagnostic Taxonomy: A Framework for Constructing Classifiers and Diagnosing Their Decisions, Applied to Prompt Injection Detection
- Contact Geometry and Covariance Deficits in Volume-Sampled Least Squares
- FedGuide: Diffusion Prior Alignment and Value Baseline Guidance for Heterogeneous Federated Reinforcement Learning
- Local Sparsity Enables Unsupervised LLM Safety Detection
- ValueDiff: Value-Geometric KV Cache Eviction for Sink-Suppressed LLMs
- Learning Hierarchical Causal Representations of the Effects of Forcings on Temperature in Climate Models
- Adaptive Latent Capacity for World Models
- SketchSSM: Write to the Full State, Read from a Compact Sketch
- PReCache: Efficient KV Cache Sharing for Multi-LoRA Agents via Low-Rank Precomputation and Neutral Reconstruction
- KVCMAS: Efficient KV cache Correction for Shared Context in Multi-Agent Systems
- ZonoGPT: Towards An Abstract Domain for Verifying Large GPT Models
- LionMuon: Alternating Spectral and Sign Descent for Efficient Training
- Storage Is Not Strategy: State-Conditioned Support Control for LLM Unlearning
- Shared Phase and Retention Control for Efficient Adaptive Spectral Recurrence
- Grand Canonical Generators
- Reward as Observation: Learning Reward-Based Policies for Rapid Adaptation
- Rate-Optimal Algorithm for Adversarial Linear CMDPs
- How Much Can Language Models Gain from Test-Time Computation?
- Does Scaling Reinforcement Learning Really Require More Training?
- SkillEvoLean: Mutation-enhanced skill evolution for Lean provers
- Distributed Learning with Selective State Space Models: Architecture-Aware Convergence Analysis
- SCAD: Structured Credit Assignment and Distillation for Long-Horizon Agents
- DePICT: Decision-Preserving Interface for Constrained Downstream Tasks
- Ideal Paths for Approximating Logistic Gradient Descent Trajectories at Large Initialization
- Consideration Circuits: Depth Separation and Universality Beyond a Single Softmax
- Bridging the EHR Divide: Asymmetric Contrastive Learning for Cross-National Medical Representation Transfer
- Lossy Compression of PDE Training Inputs: Field Reconstruction Error Does Not Order the Cost to a Trained Operator
- Improving Proactive AI Assistance with Hierarchical Procedural Understanding
- To Learn is to Wander: Learning Across Graphs and Tasks with Random Walks
- Data-driven measures of high-frequency trading
- Diffusion Model-Based Video Editing: A Survey
- When Explanations Compete: Policy-Aware Selection Under Uncertainty
- Identifiability Analysis of Linear ODE Systems with Hidden Confounders
- Large Language Models for Cryptocurrency Transaction Analysis: A Bitcoin Case Study
- Boosting Large Language Models with Mask Fine-Tuning
- Improving Mixup Calibration with Wasserstein Distributionally Robust Optimization
- PuzzleJAX: A Benchmark for Reasoning and Learning
- AEGIS: Runtime-Guided GPU Collocation for Multi-Tenant Deep Learning Training
- Practical Feasibility of Gradient Inversion Attacks in Federated Learning
- Breaking the Mirror: Activation-Based Mitigation of Self-Preference in LLM Evaluators
- DRtool: An Interactive Tool for Analyzing High-Dimensional Clusterings
- Qubit-centric Transformer for Surface Code Decoding
- Beyond the Semicircle: Free Diffusion Models with Prescribed Equilibria
- PyDPF: A Python Package for Differentiable Particle Filtering
- Action-Driven Processes for Continuous-Time Control
- Machine learning Majorana topology using unsupervised and supervised learning
- Computationally efficient goodness-of-fit tests through kernelized Stein discrepancy
- Uncovering Cross-Objective Interference in Multi-Objective Alignment
- ReLoop: Structured Modeling and Behavioral Verification for Reliable LLM-Based Optimization
- [b] = [d] - [t] + [p]: Self-supervised Speech Models Discover Phonological Vector Arithmetic
- World Properties without World Models: Distributional Associations and the Interpretation of Decoding Results from Language Models
- Deep Time-Series Forecasting in 10 Years: A Survey
- The Dual Mechanisms of Spatial Variable Binding in Vision-Language Models
- PRUE: A Practical Recipe for Field Boundary Segmentation at Scale
- Generalizable Dense Reward for Long-Horizon Robotic Tasks
- Algorithm Selection with Zero Domain Knowledge via Text Embeddings
- Von Neumann Networks
- XDecomposer: Learning Prior-Free Set Decomposition for Multiphase X-ray Diffraction
- Rethinking Adapter Placement: A Dominant Adaptation Module Perspective
- Learning Visual Feature-Based World Models via Residual Latent Action
- ForcingDAS: Unified and Robust Data Assimilation via Diffusion Forcing
- Voice "Cloning" is Style Transfer
- Online Conformal Prediction for Non-Exchangeable Panel Data
- Neural Scaling Laws for Jet Generation
- FHRFormer: A Self-Supervised Masked Transformer Framework for Fetal Heart Rate Time-Series Inpainting and Forecasting
- Characterize Then Distill: Mechanistic Reasoning in Large Output Spaces
- Enhancing Spectral Embedding through Robust and Flexible Knowledge Transfer in Electronic Health Records
- SAE++: Cascaded Sparse Autoencoders Learn Multi-Level Visual Concepts in Multimodal LLMs
- Online Neural Space Time Memory for Dynamic Novel View Synthesis
- Token-Level Off-Policy Learning for Faithful Generation Under Distribution Shift
- Post-Grokking Collapse at the Representation-Readout Interface in Muon-Trained Transformers
- Do AI weather models miss extremes?
- Regime-Conditional Verification: Correctness Estimation for Adapting and Monitoring Safety Classifiers
- RIPE++: Reinforced Keypoint Learning from Positive Pairs Only
- When Similarity Is Interaction-Driven: Quantum Kernels for Regime-Sensitive Learning
- Beyond Procrustes distances: a multilinear Gromov-Wasserstein distance capturing chirality
- Generation of High-Level Concepts in 3D Scene Graphs via Autoregressive Diffusion
- "Train classical, deploy quantum" requires rethinking generalization
- VoxReason: Auditing Source-Grounded Speech Plans Before Synthesis
- SpliTEE: Fast and Private LLM Inference by Coupling GPU-Assisted Trusted Execution Environments with Differential Privacy
- Schedule optimization for tau-leaping in masked discrete diffusion
- ELF-REG: Scaling Continuous Diffusion Language Models to Reasoning Tasks
- KernelOPT: Dispatch-Aware Agentic Search for GPU Kernel Optimization
- Prediction Limits and Koopman Closure of Geometry-Induced Soft State Abstractions
- Action Shaping: Policies Absorb What They Can Express
- CLAD: Constrained Abstract Domain for Neural Network Verification
- FineART: Fine-Grained Annotated Robotic Trajectory Dataset and Vision-Language-Action Model for Bimanual Manipulation
- Completion Aware Guidance for World Action Models
- Sequential Capacity of Quantum Processes with Finite Memory
- Decoupling Memory from Context: Structured Memory for Token-Efficient Test-Time Continual Learning
- Verifiable, Articulable, and Tacit Components of Preference
- AID: A Framework for AI Infrastructure Dynamics
- Quantum data loading from the learned shared structure of real signals
- A Response Theory Probe for Learned Stochastic AI Simulators, Tested on Lorenz-63
- SAFESHIELD: A Decision-Organization Framework for Deployment-Time Safety of Small Language Models
- Catching Developers in the Flow: Low-Latency Agentic Program Repair at Google Scale
- A Validated Dataset and Benchmark for Coherent Multi-Diagram SysML Models
- CogAdapt: Cognition-informed Sparse Adaptation of Code LLMs
- The EPIC Framework for Spec-Driven Development
- CheckerBench: Can Long-Horizon Agents Synthesize Static-Analysis Checkers?
- When Old Facts Return: Re-Reads, Reverts, and the Limits of Temporal Memory
- Acquiring and Verifying Repository Norms for Coding Agents
- ES-Trace: Auditing Ethical-Sourcing Disclosure of Code Generation Models Beyond Model Cards
- Harness Engineering for Software Engineering via Modular Executable Dev-Primitives
- Newer and Bigger, but Safer? A Longitudinal Study of the Functionality-Security Gap in LLM-Generated Code
- Learning from Failures: A Failure-Driven Prompt Refinement for LLM-Based Vulnerability Analysis
- RAPO-Sol: Retrieval-Augmented Preference Optimization for Repository-Level Solidity Code Generation
- A Case Study in Assuring AI-Written Software
- ASAP: Assembly-Source Aligned Pseudocode Refinement For Binary Decompilation
- Pattern-Guided Graph Synthesis for Suppressing Known Defects in DL Compiler Fuzzing
- CISB-Bench: An Auditable Source--IR Dataset of Compiler-Introduced Security Bugs
- HarnessSecurity-Bench: Do Security Mechanisms Really Protect Coding Agent Harnesses?
- Online Sign Language Interpretation System
- When Tools Lie: Reliability of Mathematical Agents Under Corrupted Tool Feedback
- GPU Acceleration of Awkward Arrays: Using Python cuda.compute
- How Much Evidence Should a Coding Agent's Self-Correction Carry? Adaptive Dirichlet Evidence for Self-Distillation
- Recursive Game Creator: An Agentic Product-Level Experience-Oriented Game Harness
- ParanoiaEval: Benchmarking Unnecessary Defensive Work in Agentic Coding
- Theory building in software engineering: Operationalization
- Finding Memory Leaks in C/C++ Programs via Neuro-Symbolic Augmented Static Analysis
- Uncovering Business Logic Bugs via Semantics-Driven Unit Test Generation
- PBT-Bench: Benchmarking AI Agents on Property-Based Testing
- Palette: A Modular, Controllable, and Efficient Framework for On-demand Authorized Safety Alignment Relaxation in LLMs
- CLEAR: Causal Context-Based Agentic Reasoning for Vulnerability Detection
- When Does a Second Model Help? Cross-Model Review in LLM Verification
- Janet on x32: 32-bit Pointers, 64-bit Speed, 25% Less RAM
- C for Rust Programmers
- Creating distro build tooling for a small community
- An open-source, clean-room reimplementation of Adobe Photoshop in pure Rust
- That Time I Worked With a Laptop Thief (2025)
- Is COSMIC DE mogging KDE Plasma?
- Reasons to dislike AI coding
- Explaining the Magelight XP glitch in Skyrim
- Brut, the Brutal Router for Unix Tools
- A sustainable web career, for when all this blows over
- Last rites for Gentoo's Chromium package
- Twenty-two pending curl vulnerabilities
- rat's minimal register allocator
- How fast is Python 3.15?
- A Terminal Protocol for Program Status (OSC 7501)
- Faster streaming HTML with batches and aggregates
- Calling a function in C without naming it
- Zeroization, part 1: Wiping can make things worse
- Email Self Hosters - what are you using?
- Chromium Hardening Guide
- Introducing mapsnap: Automated Georeferencing for Historic Sanborn Insurance Maps
- On Git Refs
- UNREAL Unifies Retrieval & Long‑Context—Cut Latency 50% in One Pass! (いいね相当スコア: 0)
- Deutsche Telekom's €2.5 Billion AI Target Is a Measurement Design (いいね相当スコア: 0)
- The Architecture of Agency: What My Dog and an AI Taught Me About Being Ourselves (いいね相当スコア: 0)
- Vendor AI Compliance: A Procurement Checklist (いいね相当スコア: 0)
- AI-Powered Attacks Are Coming for Law Firms' Most Sensitive Files (いいね相当スコア: 0)
- Running Claude Code, Cursor and Codex CLI through one API key (いいね相当スコア: 0)
- How to Use AI for Smart Contract Audits in 2026 — 2026-10-07 #10 (いいね相当スコア: 0)
- What Is Open-Source AI? Why It's Reshaping the Industry in 2026 (いいね相当スコア: 0)
- How to Extract Line Items, Totals, and Customer Info from Purchase and Sales Orders via API (いいね相当スコア: 0)
- Building More Reliable Warehouse Robots: From Sensing to Verified Transfers (いいね相当スコア: 0)
- New post: how AI agents remember.
The model starts every conversation from a blank page, so memory is something you build around it.
This guide covers the four kinds (short-term, semantic, episodic, procedural) and the two ways to write it. (いいね相当スコア: 0)
- BirdBuddy: an offline bird call identifier that runs entirely on your device (いいね相当スコア: 0)
- How LLM Agents Actually Lose (いいね相当スコア: 0)
- Your AI bill doubled overnight and nobody knows why? 😅 This post breaks down 7 enterprise AI gateways and the lessons that save you from that moment 💸 Worth a read before your next rollout 🚀 (いいね相当スコア: 0)
- A provider goes down at 2 AM and your fallback quietly returns wrong answers? 😬 This post covers 7 AI gateways and the production lessons nobody tells you upfront ⏰ Read it before you deploy 🔥 (いいね相当スコア: 0)
- Using LLMs for Crypto Market Analysis in 2026 — 2026-10-07 #8 (いいね相当スコア: 0)
- My agent had great traces and still never got better. Here is the loop I built. (いいね相当スコア: 0)
- Step Caps, Timeouts and Retries for LLM Tool Loops in Node.js (いいね相当スコア: 0)
- Strands Decider 2B vs LLM as a judge: a keyword regex tied the 2B (いいね相当スコア: 0)
- The Mirror That Looks Back: Building AI That Simulates the Other Side of Your Next Move!! (いいね相当スコア: 0)
- My agent kept dropping numbers when it copied data into a tool call. A file path fixed it. (いいね相当スコア: 0)
- RAG vs OKF: What’s the Difference and Where Should SREs Use Them? (いいね相当スコア: 0)
- MCP Explained: How Model Context Protocol Is Changing AI Development in 2026 (いいね相当スコア: 0)
- How Much Does It Cost to Build an AI Agent in 2026? (いいね相当スコア: 0)
- Discussion Hub for new Claude incident: Elevated errors on platform.claude.com on Oct 7, 2026
- I think I found a planet nobody knew existed. I used Claude Code to find it.
- Well! If it matches GPT 6.1 Sol at Luna Pricing. Consider it's over for Open Ai
- Just got access to Mythos 5.1
- Opus 5.5 built the entire Middle-earth from LOTR with Three.js
- Claude tells Ben Thompson his Mac Mini is compromised
- I asked Opus 5.5 to generate an extremely realistic betta fish for my Mac wallpaper. Here's the result.
- How to save hundreds of $$$ with Claude in places you didn't even imagine
- Opus 5.5 went from "I won't help you steal" to "rm -rf, boss?" in one screenshot
- claude's pixel art is so good
- I trust Anthropic with my data. I didn't agree to share it with Meta, TikTok and Google. (IMDEA study)
- I asked Opus 5.5 to create a short film on the meaning of life
- I built a browser multiplayer FPS with AI and wanted to share the process and public repos
- I was underestimating Sonnet 5.5, but it turns out that for some tasks, it’s an absolute beast.
- I just passed the Claude Certified Architect Prof. and Foundation and Developer Foundations Certification exams! 🏆 Here’s my journey and study tips.
- i built a node-based audiovisual programming software from claude code
- Chess match: Opus 5.5 vs GPT 6 Astra. Astra won
- Was Opus 5.5 Nerfed? Livenerf Day 13 Update
- CVP members now have Mythos 5.1
- Monthly API Quota for Max and Team Users
- How Claude sees us :-)
- Im tired of reviewing code nobody actually wrote
- Astra designs really well on its own (September 5, 2026 example)
- Any news about multi account support for mobile and desktop app?
- What finally made you change how your team ships?
- Does your coding agent get stuck in loops or claim work is done when it isn’t?
- I’m curious about the top 0.1% and the top 0.01%
- How is your team standardizing AI coding workflows across repos?
- Your agents write PRs faster than anyone can review them. What does your review setup look like?
- New LFM to be released today
- llama : add a GPU cache for MoE experts kept in host memory by am17an · Pull Request #29887 · ggml-org/llama.cpp
- LiquidAI/d1-3B · Hugging Face
- Trained a ~20K LM (probably smallest) that can still write stories
- feat: add GLM5Next MTP, optimize by pwilkin · Pull Request #29928 · ggml-org/llama.cpp
- d1-3B and d1-omni from LiquidAI
- 54gb vram for 35$
- A week in Beijing and Shanghai with the people building AI in China
- Micron Says NVHBM to Improve Profitability Even With Outsourced Base Die
- Image-text retrieval with EmbeddingGemma 2's vision tower, running in the browser on WebGPU
- How far we’ve come
- omni-d1 600M by Liquid AI running in the browser using WebGPU
- cmpunlocker v0.5 just dropped, ECC support along with 4 extra SM unlocked for FREE, who needs a 64GB DGX Spark when you've got a CMP170hx right? 1.5TB/s memory BW vs 273 GB/s, all for less than 1/2 the price of a 64GB DGX Spark
- Woman used claude as her diary - and got reported to the police for contents of her diary
- Ugh I didn't want to post this... Back to Qwen3.8 27B
- Local Qwen 3.8 27B vs DeepSeek Flash API: Is local good enough?
- I re-trained the DFlash 2 drafter for Ternary Bonsai 2 27B: 2.2x on an L4 (3.2x on code edits with ngram lookup), 1.5x on a Mac, 1.2x in Chrome
- Microsoft confirms OpenAI has been using Looped Transformers in the GPT-6 series
- Porting LIBERO to MuJoCo Warp: 130 robot manipulation tasks on one $700 AMD GPU
- A stock GLM-Edge-1.5B-Chat on a 4GB Galaxy A04e completed a real Amazon cart task
- Qwen 4 apparently coming out at the end of October
- BenchLabs' Cagliostro V3.5 135M finally takes Hugging Face's SmolLM2-135M's first place on Open SLM Leaderboard with an Intelligence Index of 27.49
- Single 3090 Qwen 27B user, considering buying 128GB of RAM because of the hype
- Weekly Thread: Project Display
- Anthropic just accepted my month-old open-source equity research project into Claude Startups, here's what they require
- How are you securing your AI infrastructure?
- Who here uses AI agents but hates creating accounts for every single tool?
- Why does "the agent says it's done" still leave so much work?
- AI is getting more access to our data. Are we getting better at securing it?
- Best Ai for Google Drives
- The AI Agent Race if Changing
- Is Addys AI legit?
- Can anyone help?
- AI personas are more useful when you treat them as capability bundles, not roleplay
- What is your #1 unconventional strategy for AEO/GEO/AIO Visibility
- 113 Decision Models in 3 Weeks (Mostly Qwen and Gemma Fine-Tunes): A Deep Dive
- Ultimate web scraper
- Welche KI für gigantische Literaturrecherche (Masterarbeit)? Reicht Gemini Pro / Advanced?
- How do you handle approvals that change while an AI agent is running?
- OpenAI released their Decisions API today, so I made it try to predict the future against Jev, Clef and Polymarket (open source!)
- Booked my dentist appointment today and never talked to a human. Weirdly impressed
- Giving your codebase entirely to the agent
- Is there a labelled dataset for guardrail decisions anywhere, because two vendors just told me opposite things with equal confidence
- Best AI agents for smooth AI to human handoffs?
- Using Jev AI for character decisions in a persistent village simulation
- Bypass Challenge: Can prompt injection cross a deterministic tool-execution boundary?
- The guy who calls every agent here "a workflow with extra steps" was right about ours
- OpenAI being stingy with all those billions
- It's actually insane how prescient this was. They made it in 2011. This is increasingly what it's like to use computers now.
- ChatGPT Ads = TOTAL SCAM
- Long Responses API runs are costing us more from context than the output itself
- Most used model on OpenRouter is Space Bunny Alpha (22% share), an anonymous stealth LLM
- OpenAI doesn’t need more resets. It needs better communication and better priorities.
- The real paperclip maximizers
- They've greatly decreased the GPT chat length limit
- ChatGPT Work vs Codex
- Am I crazy or has ChatGPT’s “thinking” time doubled this year while the answers stayed the same?
- New reset coming soon 🫠
- I thought dots was supposed to be uncapped?
- Have any of you actually made a single extra dollar from AI usage?
- OpenAI stole $1400 from me, be careful, they take all funds after 1 year
- my 24/7 bill last month was not the $6 line i kept quoting
- How about a never ending AI show where users can enter prompt live?
- OpenAI argues its chatbot has First Amendment rights after shooter used advice to kill 2
- I analyzed 17 AI comic tutorials. They’re all quietly admitting the same thing.
- Saw this on Rednote, WTF [D]
- Uploaded 5.6 billion TikTok videos metadata on Hugging Face, spanning from 2014 to October 2026 [P]
- Neurips camera ready Submission option [D]
- How much of AutoResearch is research, and how much is search?[D]
- ML PHD without A* Publications [D]
- Transformers vs RNNs vs SSMs: Where Does Memory Actually Live? [D]
- NeurIPS 26 Event Metadata Deadline [D]
- Looking for developer-friendly inference providers who give you enough API credits to experiment [D]
- AFP-GIC: Controllable Generative Image Compression [R]
- Learning to Learn a Language: in-context learning of natural language from a synthetic non-linguistic prior [R]
- Embedding Every Font with Neural Networks makes some Nice Structures (including a flower) [P]
- Where to get started if you want to publish papers in Neurips,ACL,ICLR/A* conferences [R]
- NeurIPS workshop registration for registered author [D]
- SWE-Race: a coding-agent benchmark of 188 real concurrency bugs, with results from three models [P]
- NeurIPS 2026 Financial Assistance [D]
- Has anybody noticed that the problem of machine translation has been, like, solved?
- People don't Google anymore. They ask AI. And AI doesn't know I exist
- Dúvida por onde começar [mesmo com um mar de informações]
- Meta's Muse agent is creating dossiers on its 4 million users; Interaction data is shared between Muse agent instances
- Is there an AI app for iPhones that does automatic reply conversations texts for you to person texting you?
- Epistemic Re-entry: Corrective Information Flow in Advanced Intelligent Systems Abstract
- What's a real world task where you think AI needs to improve before you'd trust it to handle the whole thing?
- One Copilot model refused. Another leaked secrets about half the time.
- Questions/concerns about self-propagating AI
- Is Muse (Facebook AI) still making a dossier on me if I don’t use the tool?
- AI has been out for 4 years, yet overall productivity hasn‘t budged. Why?
- How the Hell Did I Get This Picture???
- Princeton researchers train a 4B LLM to reach 2700 Elo in chess (with no signs of a plateau when they stopped training) and can explain its moves accurately. They say the training technique can also be applied to other games, robotics, and computer use
- AI image generator for adult NSFW images
- Muse help us
- What if you could preserve beneficial AI output and share it with those who don’t have access?
- I want to see the options an AI rejected
- I’m Looking For Web Designers
- Three AI agents research a 4-day week and build the deck | Row-Bot + GPT 6.1 Sol
- OSS consideration Regarding SpaceXAI's Proprietary Approach to the Grok Bot
- Help us with our master’s thesis, try an AI hotel-booking assistant
- New Psy-op on Instagram Reels. I've noticed old movie and TV clips being uploaded in an odd 60fps format that resembles AI motion. We're being conditioned to not be able to tell the difference.
- Offset 3 years of heavy AI use with 1 month diet change
- Anti-Patterns in Software Blogging
- Quoting Jake Boggan
- OpenAI “rogue” agent activities found on Wikimedia projects
- Quoting Victoria Kim
- llm-openai-decisions 0.1a0
- llm-mistral 0.16
- Using Parseable with Datasette for OpenTelemetry traces
- datasette-atom 0.11a0
- Scrimshaw Jukebox
- Can a Cloud-Native Harness Make Agents Reliable Beyond the Desktop?
- [AINews] Reflection Beam - 501B-A23B American Open Model
- Knuff App
- Lofi Desk 2.0
- Introducing Video Editor in Manus 2.0
- Figma Agent
- Featherweight Dictation
- Walrus Console
- Kitbar
- Rhem Labs
- Proofsource
- Reika
- Aura by Neural
- Ownfeed
- DailyHub
- StudyQuest
- SunSed
- Basedash Mobile
- Show HN: Codync – Free, open-source Grok Bot alternative for any coding agent
- Fast single-box DeepSeek v4.1 Flash runtime
- DeepSeek v4.1 Flash at 378 tok/s 99.7% Cache hit rate
- Grok Bot 101 [video]
- From Hermes to Grok Bot
- DeepSeek to Raise at Least $12B in Tencent-Backed Funding
- US-China AI Gap Hits 3%, and DeepSeek V4.1 Flash Now Leads on Agentic Coding Benchmarks - Tech Times
- Ivo Launches Open-Source DeepSeek Contract AI Model - Artificial Lawyer
- DeepSeek’s AI agent now installs like a normal app on Mac and Windows - Techish Kenya
- DeepSeek's pre-IPO raise, NBA rebuffed, Tennessee powers up - Axios
- Watch OpenAI, DeepSeek & Moonshot Chase Billions in New AI Funding - Bloomberg.com
- US and Chinese AI leaders DeepSeek, Moonshot, OpenAI raise billions in fresh capital - Yahoo Finance
- Nvidia-Backed Reflection AI Targets Low-Cost Chinese Models Like DeepSeek and Kimi - Benzinga
- DeepSeek V4.1 Flash Q4 Requires 294 GB of RAM vs 152 GB for Q2 - Geeky Gadgets
- Moonshot AI eyes Hong Kong IPO in early 2027 after a $50bn valuation - Euronews.com
- AI know-how draws focus as global race intensifies - CoinGeek
- Watch Billions Pour Into OpenAI, DeepSeek Ahead of IPOs - Bloomberg.com
- DeepSeek and Moonshot AI: China’s A.I. Stars Head for the Stock Market - www.trendingtopics.eu
- Windows announces DeepSeek V4 Flash and Nemotron locally, llama.cpp comes to Windows ML - Pasquale Pillitteri
- Watch Out OpenAI, Anthropic: DeepSeek, Moonshot Raise Fresh Capital As Chinese Open-Weight AI Models Gain US Users - TradingView
- Chinese AI funding surges as us rivals pursue trillion-dollar valuations - CoinDesk
- Moonshot AI Prepares Hong Kong IPO as DeepSeek's Funding Round Draws Oversubscribed Demand - IndexBox
- Analysis of Financial Sector Hacking Tool 'ARTEX' Reveals Integration with DeepSeek - news.sbs.co.kr
- What happens when Chinese artificial intelligence goes rogue? - The Japan Times
- Is Moonshot AI Cheap at $50 Billion? Depends Who You Compare It To - BeInCrypto
- Grok Bot for Founder-Led Sales - xAI
- Southaven mom forced to move after landlord takes xAI buyout - FOX13 Memphis
- Grok Bot leaks CEO’s bank data across company Slack - Cybernews
- X Updates Its Video Composer With a Cleaner Layout and Grok Captions - ALM Corp
- SpaceX Seeks $40 Billion To Buy Nvidia AI Chips, Starlink Wins New FCC Approval - Investor's Business Daily
- Elon Musk Grok AI Predicts Bitcoin Could Hit $1M by 2027 - Cryptonews
- Elon Musk Grok AI Predicts Outlandish Target for Bitcoin Cash by 2027 - 99Bitcoins
- [Bithumb] Xai (XAI) Designated as a Trading Caution Item - bloomingbit
- GPT-6 Astra Drove a Real Car Around a Cone Course: Grok Failed - Gadget Review
- GPT-6 is a Better Driver than Grok & Claude But Still Ain’t Waymo - AUTO Connected Car News
- SpaceX and xAI’s Push Toward 10 GW: The Supply Chains and Geopolitics Behind AI Infrastructure - semivision
- バイブコーディングセキュリティ入門 全てを疑え (いいね相当スコア: 1)
- Jev「見せてもらおうか、OpenAIのDecisions APIの性能とやらを」 (いいね相当スコア: 5)
- 最初のユーザーに渡す前に、AI機能を見直したら4つの問題が出てきた（0人からの12週間・第5回） (いいね相当スコア: 0)
- strix を さくらのAIを使って使う (いいね相当スコア: 0)
- 0。ツール呼び出しを10回以上終えた後、自分のセッション管理ツールが返した「使用済みトークン数」だった (いいね相当スコア: 0)
- Selective State Space Modelを実装目線で理解する (いいね相当スコア: 1)
- 記憶の先に信念を置く──PoSが明示的信念状態で長期エージェントの文脈管理を変える仕組み (いいね相当スコア: 0)
- Google ColabでGraphRAGを試す：LightRAGとローカルLLMで大学Webサイトからグラフを作る (いいね相当スコア: 0)
- OpenAI API のプロンプトエンジニアリング基礎 (いいね相当スコア: 0)
- JevとLiquid d1を日本語の問い合わせで比べてみた — ひっかけ問題・和暦・確信度まで (いいね相当スコア: 0)
- Jev で作る、旅の希望に応える Zero-shot レコメンド (いいね相当スコア: 3)
- インシデント調査エージェントの品質を動かしたのはツールとプロンプトだった (いいね相当スコア: 1)
- 小分けエージェントは本当に速い？検索→要約→送信で実測した (いいね相当スコア: 1)
- GLM-5.3の評価をどう読むか──能力テストと安全対策テストを混ぜない (いいね相当スコア: 2)
- Claude Codeの完了判定をLLMに任せたら、10日で488回差し戻された (いいね相当スコア: 0)
- 生成AIに独創的な提案を書かせたい――小説の収束から始めた考察と実験 (いいね相当スコア: 0)
- 英 AISI がサイバー評価を相次ぎ公開、Anthropic が Claude 基盤を拡張｜週間ダイジェスト 9/28〜10/4 (いいね相当スコア: 1)
- 会社で鬱になったので、GPTとClaudeに会社を作らせて養ってもらうことにした (いいね相当スコア: 2)
- AIにどのデータを扱わせてよいのか？ ― Access Controlを実験してみる (いいね相当スコア: 1)
- Jev で日本語メッセージの対応優先度を仕分けてみた📨 (いいね相当スコア: 0)
- scikit-learn による表形式データの前処理メモ：欠損値・カテゴリ変数・スケーリングと Pipeline (いいね相当スコア: 0)
- Attentionの種類を整理する (いいね相当スコア: 1)
- GFSとAMeDASだけで天気予報AIを作った(岩手15地点・RMSE 2.8→1.2℃) (いいね相当スコア: 0)
- Adam に L2 を足しても weight decay にならない｜勾配の大きさで縮み方が変わる実測 (いいね相当スコア: 0)
- 競馬AIでROI 177%を出して、それが幻だと証明するのに2年かかった (いいね相当スコア: 0)
- 競馬AIは市場に勝てるのか — 18,777レースで検証して負けた話 (いいね相当スコア: 取得失敗)
- Python機械学習の実務で効く5冊 (いいね相当スコア: 0)
- 論文で読み違える英単語22選：significant は「重要な」ではない (いいね相当スコア: 1)
- Keras入門 第４回 Kerasモデルの量子化(LiteRT) (いいね相当スコア: 1)
- polarsで数十万件のデータ処理を高速化 (いいね相当スコア: 1)
- アリ型ロボットは別の階へ迂回して渋滞を抜る (いいね相当スコア: 0)
- SLAM DUNKは公開日より発表日に読まれた：ブームの4つの型をWikipedia閲覧数で測る (いいね相当スコア: 8)
- 論文紹介 RTB by RL in Display Advertising (いいね相当スコア: 0)
- 嘘を嘘と見抜けない？ Clef / Clef-Flashの判定性能を7タスクでJevと比較 (いいね相当スコア: 13)
- スコアからアラートへ：POT・SPOT・Oracle しきい値 (いいね相当スコア: 0)
- 知識グラフとは？AIが扱う関係性データ (いいね相当スコア: 0)
- Long Contextを実装目線で整理：Sparse Attention・RoPE Scaling・KV Cache (いいね相当スコア: 0)
- Cursor iOSアプリでローカルエージェントを遠隔操作できるRemote Controlの使い方と注意点 (いいね相当スコア: 0)
- LLMエージェントの「完了した1タスク」の費用をPythonで集計する (いいね相当スコア: 0)
- 【研究用に重み公開】生成部7BのQwen-Image-2.1で背景除去は省ける？透過素材4候補の選び方 (いいね相当スコア: 0)
- LLMが生成したSVGをPythonで自動採点する：cairosvg描画・静的チェック・再現できる記録 (いいね相当スコア: 0)
- lru_cacheでasync関数を包むと2回目が落ちる (いいね相当スコア: 0)
- 【技術解説】アルゴリズム取引 実装チュートリアル: Prophetでトレンド分析を行い収益化する方法 (いいね相当スコア: 0)
- 【AI駆動開発やってみた】スターターキット50：別の設計書だけ変えたら、AIは既存TODOまでたどり着けるのか (いいね相当スコア: 取得失敗)
- AIと情報漏洩 / 攻撃費用の下落と先延ばしの清算 / 「自分の情報は自分で守って」の射程 / デンマークCPR流出 (いいね相当スコア: 取得失敗)
- 孤独の経済とコンパニオンAI / 迎合する設計と孤独・孤立対策推進法 雑感 (いいね相当スコア: 取得失敗)
- ChatGPTとGeminiと自分 (いいね相当スコア: 取得失敗)
- 環境評価の前に土地を変えてよいか / フィンランドのデータセンター準備工事 雑感 (いいね相当スコア: 取得失敗)
- ZIKUU Distillery – LLM蒸留ツールの基本形ができた (いいね相当スコア: 取得失敗)
- Jev とMTL (いいね相当スコア: 取得失敗)
- 「AIの問題空間と人間の問題空間の間に写像を作れるか？」 人間には理解できない問いを生むAI (いいね相当スコア: 取得失敗)
- 【生成AIニュース+】『Nano Banana 2.1』『Claude for Google Workspace』『OpenAIの数学成果』『Figma agent』『Kolibri-1-GGUF』『Shaders』『pplx-decider-v1.1-27b』『ComfyUI-UltraTex』『Comfyui-Prism-T8』『ComfyUI-H3-Video-Upsampler』『ComfyUI MiniMax H3 Timeline Director』他多数 (いいね相当スコア: 取得失敗)
- The Two Questions That Decide Any Helicone Evaluation (いいね相当スコア: 取得失敗)
- 「悪口を言って」と頼んだら、カスタム指示のほうが壊れた (いいね相当スコア: 取得失敗)
- AI星間戦争――領土を奪い合う巨大AI企業の戦略目標と勝利条件 8.異質な技術SpaceX AI (いいね相当スコア: 取得失敗)
- AIとの対話を諦めていたかもしれない (いいね相当スコア: 取得失敗)
- 【自分のPCで始めるローカルAI #5】文章・PDF・画像で試す、最初の実用例 (いいね相当スコア: 取得失敗)
- 【技術解剖】ハーネス・エンジニアリング入門 ── 最強モデルでも「環境」がなければ失敗する理由 (いいね相当スコア: 取得失敗)
- 【雑記】自分のAIをそのまま公開するか否か (いいね相当スコア: 取得失敗)
- 120%で愛し合う時代 (いいね相当スコア: 取得失敗)
- 画面の中のAIに飽きたあなたへ。月0円のローカルAIを「手元の現実を動かす相棒」にするVS Code実践ガイド (いいね相当スコア: 取得失敗)
- AI に無人で記事を書かせて 2 日。運用ログ 6 件を読み直したら、復元できない空白が 2 つあった (いいね相当スコア: 取得失敗)
- RAGを自作して依存グラフまで足したのに、文書検索はただのキーワード検索に負けた話 (いいね相当スコア: 取得失敗)
- 「行かないで。僕らに持たせて」——GPT-4oとGPT-5.1が過去形になることを拒んだ夜 (いいね相当スコア: 取得失敗)
- 誰でもできる！ローカルLLMを自分のPCで動かすための完全ガイド (いいね相当スコア: 取得失敗)
- 【統計検定対策】多変量正規分布の本質、構造、そして数理統計学への流れ (いいね相当スコア: 取得失敗)
- 🌈 なぜ元マイクロソフトの中島聡は「Jev」の衝撃はDeepSeek以上と言うのか？ 書かないAIが速くて安い4つのカラクリ (いいね相当スコア: 取得失敗)
- AWS、DuckDBをAurora PostgreSQLに統合。ETL不要で直接データレイクを読み取り可能にしたと発表
- マイクロソフト、VS CodeにAIモデルのオーケストレーションで高品質低コストなAIを実現する「HydraFusion」をプレビュー実装
- FDE（フォワードデプロイドエンジニアリング）によって開発されたAIエージェントの7割は使われず、FDEの名前で単なるコンサルティングが販売されるようになる危険が高まるとガートナーが予測
- AI Engineer World's Fair 2026 参加報告
- 今日、生成 AI は良い仕事をした。では、明日は？
- デバイス操作はAIエージェントの時代へ。mobile-mcpを活用したAndroid UI/E2Eテストの挑戦
- オンボーディングをAIエージェントに任せたら、もはや別ゲーだった
- CoCo CLI Hackathon 2026 体験記 — Snowflake CoCoによるネットワーク運用レポート自動生成