AI News Digest 2026-10-20
台本で使った記事
特集
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
- MiniMax M3.1 Flash shows China's AI price race has no floor in sight
- DeepSeek Opens Its In-House Search Engine To Developers - Search Engine Journal
- DeepSeek teams up with Huawei to challenge Nvidia - Yahoo Finance Singapore
- DeepSeek V4.1 Flash 1017: what DeepSeek says changed, and swapping it in without breaking anyone's saved model (いいね相当スコア: 0)
- DHS Bans DeepSeek From Federal IT Systems [2026] - https://tech-insider.org/
- Mistral AI Releases Timestral: A 209M Forecasting Model Family That Outperforms Google’s TimesFM-3
- NVIDIA Run:ai Is Now Open Source as Part of NVIDIA DSX AI Factory Platform
- Google Open-Sources Harness2, a Recursive Agent Harnessing Framework
- Disney agrees to hire the majority of Character AI’s workforce
- OpenCode hack exposes API keys amid AI distillation attacks - BleepingComputer
- Grok Can Pay for Itself: What Musk's Claim Actually Means - BASENOR - Tesla Accessories
- inclusionAI/Ming-3.0-audio-flash · Hugging Face
- Mercury 3: Diffusion at scale
- I no longer care to work on any Ruby gems so I've archived all my repositories
- Degraded availability for GitHub Actions, GitHub Pages and Webhooks
- CERN and Mistral sign a statement of intent to establish a strategic partnership
- Wine ported to OpenBSD
- Alternatives to animal testing are finally taking off
- Let's Encrypt is adding two additional remote perspectives for domain validation
- Investing in the public benefit
- v2.1.303
- How to continue a GitHub Copilot app session in VS Code
- Introducing Clairic, a new System One class of extraction models
- Beyond the patch: Protecting your NetScaler gateway against future vulnerabilities
- GenRecs Tokenization: Transforming Ads Recommendation at Meta with Stable Model Vocabulary
- The Thin Orchestrator: How We Made Our AI Agents More Reliable by Giving the Router Less to Do
- Inside Prometheus: Meta’s Journey to a Gigawatt-Scale Supercluster
- Dynamics of Sentiment and Attitudes Toward AI Among Developers
- JetBrains .NET Day Online 2026: Watch the Recordings
- Can you bake Maven into a neural network?
- Accelerating Mamba-2 with High-Performance Forward Pass Kernels in NVIDIA CUDA Tile
- Making the internet better for AI agents might make it worse for everyone else
- “We were looking at it wrong”: The explosive claim that’s upended the AI world
- AI researchers stranded Claude on a virtual desert island. It started to “play.”
- AI game clones are making indie developers afraid to show their work
- Anthropic gives outside AI watchdog “employee-like” access
- Tolan snags $30M for its AI companion that helps users navigate life
- Replika’s founder built the AI assistant she pitched at Disrupt 12 years ago
- Tenstorrent hits $6.2B valuation to scale open-weight AI chips
- People are using AI to make their smart TVs dumb again
- Nvidia launches new open models to run agents in the enterprise and on PCs
- Laplace Intelligence red-teams robots and self-driving cars before they hit the real world
- Anthropic says it’s committing 5% of its shares to a ‘public benefit fund’
- JB Pritzker thinks AI could help us live to 100 — or threaten us like nukes
- A hacker broke into John Deere and SharkNinja with an army of AI agents
- Democrats think cracking down on AI could give them an edge
- A woman was arrested in Philly for blocking a delivery robot
- Article: From Microservices to Agents: the Next Evolution of Distributed Systems
- Podcast: The Future of Web Apps: Will AI Agents Generate UIs on the Fly?
- EmbeddingGemma 2 Expands beyond Text to Support Code, Images, Video, and Audio
- Edge AI for Smart Cities: From Insight to Cognitive Action
- Kling AI drops version 4.0 with stereo audio, multilingual lip sync, and a lighter Flash variant
- Two German courts now hold Google liable for what its AI overviews say
- Stripe says AI agent purchases are adding up, but it won't share the numbers
- The neocloud shakeout is coming, and Nvidia has the most to lose
- Anthropic plans fund for workers threatened by AI, Musk counters with SpaceX shares for children
- Anthropic has nearly stopped hiring junior staff because AI agents now work six times as many hours
- Four Gemini 4 Argon agents beat Minecraft without any human help
- Anthropic consults law firms to set fair negotiation norms before AI agents start cutting deals
- AI agents have bad taste in long-term decisions and a new benchmark proves it
- pnpm、発行先と違うレジストリへCIトークンを送らないよう変更
- GitHub Actions、依存をコミット固定するロックファイルを公開
- Vercel、常駐ワーカーDaemonsのパブリックベータを開始
- React Native 0.88、イベント配送をWeb標準の経路に変更
- OpenAI、GPT-4oとGPT-4.1を2027年1月14日にAPIから削除
- 2026年10月12〜19日の週：モデル更新、実行境界、画像配信、Lambda
- OpenBMB、MiniCPM-V 4.7を公開 シリーズ初のMoE 35B-A3BがiPhone 17 Pro Maxで約3GB・13トークン/秒超、専門家の先読みで性能を落とさず
- vLLM、Kimi K3のGB300向けデコード経路をNVIDIAと高速化 CuTe DSLで演算と通信を融合し最大1.9倍、エージェント支援で開発
- Apple、App Review Guidelinesに2.5.19を新設 隠し・休眠・未記載の機能を明確に禁止、難読化での隠蔽はアカウント停止の対象に
- Plex Media Serverの4脆弱性を連鎖させる未認証RCE「Plex2Shell」が公開 PoC公開後に旧版への攻撃報告、1.43.3.10861以降へ更新を
- TinyFish Makes Search Free: What It Really Costs an AI Agent to Read the Web
- How to Make Zero-Training Tabular Predictions with NVIDIA Structured Data Models
- Meet Voxel51 3.0: A Unified Multimodal Data Platform for Physical AI
- When is a large language model a good calibration agent? Physics-labeled prompting for budgeted engineering simulator calibration
- Trust but Verify: Designing Region-Inhomogeneous Forward Dynamics for Discrete Diffusion
- Binding-Specific Relearning Savings for Test-Time Continual Learning
- CGDT-CESN: Correlation-Guided and Dynamics-Trained Clustered Echo State Network for Multivariate Time-Series Prediction
- BOLT: Multi-Bottleneck Bound-Guided Learning for Flexible Job-Shop Scheduling with Time Lags and Resource-Coupled Transport
- TRACE: Transformer Residual Activation Codebook Encoding for Communication-efficient Distributed Edge Inference
- Generative adversarial networks with Monte Carlo dropout uncertainty quantification for structural health monitoring: Fibre Bragg grating data augmentation comparison in Hawk T1A aircraft
- CF-NODE: Counterfactual Spatial Node Intervention for Test-Time Adaptation Under Domain Shift, Evaluated Across Five Diabetic Retinopathy Datasets
- Rank Measures of Fine-Tuning Updates Need a Noise Baseline
- Sparse Autoencoders Can Induce Local Geometry for Mitigating Catastrophic Forgetting
- Lean4AgenticRL: Advancing Agentic RL with Verifiable Multi-level Process Rewards
- Progressive Misalignment Can Emerge from Benign Recursive Self-Distillation
- Generative Value Classifier
- Predicting Privileged Information Effectiveness for On-Policy Self-Distillation
- Learning smooth dynamics via Kinematic Interpolants
- Don't Retrain, Remix: Adaptation of Web Agents using Mixture Hypernetworks
- On the SPOT: Enhancing Speculative Decoding for Diffusion Language Models via Commit-Order Prediction
- Deterministic Bounds on Accuracy and Stability of Metric-Space Regression and Classification
- Coverage Cliffs in Off-Policy Quantile Certification
- Structure-Aware Spatial Kernels for Gaussian Process Distributed Port-Hamiltonian Systems
- Noise-Robust Forced Alignment for Spontaneous Speech
- Initiating Removal from Operational Symptoms in Federated Vehicular Networks
- Flat Loss, Evolving Features: Grokking in Modular Addition
- Diagnosing GNN Failure Modes with Layer-Wise Probe Entropy
- Can LLMs predict the outcomes of machine learning experiments?
- Rarity of Singletons Matters & Supplementary Materials
- Noise-shaped Sampling and Annealing for Scalable On-the-edge Training of Binary and Ternary AI Models
- Understanding Stability of Outer Nesterov Momentum on Stochastic Pseudo-Gradients
- Network-Wide Freeway Speed Nowcasting and Forecasting from Sparse Fixed Sensors with a Probe-Supervised Graph Convolutional Recurrent Network
- Alert-List Concordance for Evaluating Imputation Strategies in ICU Sepsis Prediction
- The Ephemeral Cost of Switching Optimizers Across LLM Training Stages
- Edit Discovery in Human-AI Interactions
- AlphaBet: Reinforcement Learning for LLM Event Forecasting Agents
- Chance-Constrained Personalized LLM Routing with Online Preference Learning
- Normalizing Flow Maps
- Spectral Limits of Polynomial Graph Filters on the Long-Range Ising Model Benchmark
- OrderFlowDiT: Transport-Regularized Diffusion Transformer for Probabilistic Limit-Order Supply Forecasting
- Native Geometry vs. Linear Recoverability: How Alignment Changes Platonic Scaling
- Refusal Is Not Robustness: Measuring Harmful Reachability in Autoregressive Language Models
- Acquisition Function-Guided Generation for Bayesian Optimization over Discrete Spaces
- Reusing Calibration Gains across Model Representations
- Next Token or Vector? Learning How Far to Decode
- Ori: Drift-Aware Streaming $k$-PCA for Very Low-Rank LLM Training
- Partial-Revelation Panel Forecasting: Predictive Information Gain, Allocation, and Coarsening-Consistent Operators
- Benchmarking and Enhancing Agentic Causal Discovery
- Learning to Be Guided: Shaping Latent Geometry for Classifier-Free Guidance
- Batched Contextual Reinforcement II: Does RLVR Need Long Rollout Windows?
- What Accuracy Hides In Learned Transformer Programs
- A Feature Selection Layer Approach for Multi-Class Classification on Neural Networks
- Scalable Levenberg-Marquardt Training Framework for Deep Learning
- Large Diffusion Models Need More Than EMA
- On the tradeoff between robustness and geometric accuracy in adversarial training
- Towards Hyperparameter Transfer for Differentially Private Optimization
- TALLOP: Joint Optimization of Local Training and Communication Compression for Time-Constrained Federated Learning
- A Label Noise Robust Algorithm for Graph Data using Neighborhood Label Distribution
- A Minimal ReLU Classifier on Non-Linearly Separable Data: Convergence and Implicit Bias at and beyond the Edge of Stability
- "Deep lifts": Neural networks can generalize even with the dumbest optimizer
- Towards Looped Models Done Right, Part I: Topology, Input Injection, Recurrent-State Design
- GR-OPD: Group-Relative Sequence Credit for On-Policy Distillation
- Path to Prior-Data Fitted Relational Models via Learned Latent Multi-table Representations
- Where Latents Belong and How They Flow: Joint Identification of Structural Roles
- Towards a GraphNARX model for structural dynamics
- Domain-specific language alignment of multimodal embeddings towards explainable predictions
- LED: Improving Masked Diffusion Language Models by Lookahead Editing
- MetaCurate: Evolving Curation Filters for High-Quality Agentic Coding Data
- Local proxies for memory-efficient training
- Grounded Credit Assignment in Reinforcement Learning for Multi-Turn Search Agents
- Nearly Linear Private Learning of Littlestone Classes
- Structural Affinity: An Algorithm for Coherent Cage-Based Shape Design
- Tree-Aware Posterior Distillation for Long-Horizon Agentic Reinforcement Learning
- A Stronger Teacher Within: Strengthening MoE Models via On-Policy Bidirectional Distillation
- LangNCO: Compiling Natural-Language Constraints into Inference-Time Interventions for Neural Combinatorial Optimization
- Non-Parametric Optimization for Scalable Learning in Stochastic Decision Problems
- Process Reward Modeling with Reliability- and Conservatism-Aware Bayesian Belief
- Response-Informed Value Estimation for Cost-Aware Multi-Turn LLM Routing
- Anisotropic Representation Dynamics in Transformers
- FLARE: Physics-Grounded Evaluation of Language-Model Agents for Feeder Restoration with Degraded Observations
- Sample, Mix, or Verify? Allocating Test-Time Compute under Imperfect Verifiers
- ADT: All-Atom Diffusion Tokenizer for 3D Molecular Structures
- Beyond Marginal Influence: Interaction-Aware Memory Selection for LLM Agents
- DeDiCa: A Scalable Deconfounding Discrete Causal Generative Model
- Data-Efficient Classifier Pretraining Through Diffusion Models
- Untrained CNN and coefficient-space regularization for joint reconstruction in photoacoustic tomography
- Peak-Memory Schedules You Can Trust: Exact Realization, Certified Optimality, and the Limits of Recomputation for Inference
- HiPredict: Predicting the missing HiRISE RED4 spectral channel with machine learning
- Minimax-optimal Scaling Laws in Ridge Regression via Transfer Learning
- SETAREH Transform: A Sorting-Assisted Haar-Wavelet Sparsification Method for Approximate Inner-Product Computation
- How Finite-Rank Bottlenecks Shape the Low-Rank Adaptation Landscape
- RTSM: Adapting Recursive Models to Multivariate Time-Series Classification
- Forecasting Voltage Time Series under Anomalies Using a WGAN Model With Lipschitz Control and Progressive Kernel Critic Strategy
- Contracting with an Agent under Model Mismatch
- Information Estimation with Discrete Diffusion
- Flash RNN-T: Memory-Efficient Joint Network and Transducer Loss Computation
- Residual Stochastic Concept Bottleneck Models
- Bridging Compressed Communication and Local Steps via Spectral Method in Federated Learning
- RECREG: Recovery-based Shielding for Regular Safety Properties in Reinforcement Learning
- Regularized free deconvolution for spectral repair of resampling distortion in high-dimensional particle filters
- Predictable Scaling: Efficient Model Exploration via Scaling-Law Prediction
- A Topological Signature of Learning under Noisy Labels
- Uncertainty-aware AI-enabled framework for predicting hydrogen dispersion in confined spaces to enhance operator situational awareness
- Out of Support, Not Out of Reach: In-Context Learning with Unsupported Instructions and Composed Tasks
- Beyond Expected Values: Risk-Sensitive Planning with Distributional Monte-Carlo Tree Search
- BMuon: Budgeting Muon's Spectral Geometry for Structured Compressibility
- Boundary Detection for High-Dimensional Point Clouds
- When Magnitude Suffices: Reconstructing Time-Domain-Gated Microwave Signals Without Phase on a Thoracic Phantom
- How Many, but How Sure? Class Competition as a Confidence Proxy in Multi-Class Density-Based Counting
- The Sign Code: The Hidden Binary Nature of Deep Networks
- ThetreStageAI: Multi-View Retrieval-Augmented Generation for Structured Bengali Theatre Production Planning
- GRAFT-EEG: Gauge-Aware Physical Field Transport for Cross-Dataset Motor-Imagery EEG
- Deconstructing Multi-Task Active Learning: The Paradox of Gradient Conflict and Orthogonal Decomposition
- Deep LLM layers expose passage-level context in EEG during silent reading
- No-Regret Language Generation
- JAXtari: High-Throughput and Easy-to-Modify Arcade Learning Environment
- TriRouter: Sparse-to-Dense Electrode-Space Routing for Robust EEG Motor Imagery Decoding
- Equivalence testing and run-count design for stochastic machine-learning experiments
- Blitz: Accelerating the Diffusion Modules of AlphaFold3 and Boltz-2
- Tool Recommendation for Multi-Step Agents: Finding Every Tool a Task Needs
- Neural Expansion: A Unified Mechanism for How Deep Neural Networks Generalize
- Escaping Parameter Space: Tight Generalization Bounds via Representation Quality
- LagNet: Lag-Structured Operators for Long-Term Time Series Forecasting
- Universal Semantic Enhancement for Weakly Supervised Anomaly Detection
- Logic-Guided Graph Denoising Diffusion Models
- Explainable Machine Learning Analysis of Temperature Effects on Mechanical Strength of Fiber-Reinforced High-Performance Concrete
- Learning Personalized Memory Gating from Delayed One-Sided Feedback
- CocoRNA: Collective RNA Design with Cooperative Multi-agent Reinforcement Learning
- Waiting Buys Exactly Two Moments: Endogenous Identification in Prediction-Driven Allocation
- Property Estimation in Geotechnical Databases Using Labeled Random Finite Sets
- How language model agents reshape price competition
- Encoded but Exposed: Language Embeddings can Reveal More Than They Should in Recommender Systems
- Physics-Transfer Learning: A Physics-Constrained Theory for Zero-Shot Generalization Along the Ladder of Complexity
- Memory Bandwidth, Not FLOPs: Profiling and Accelerating Query-Based Segmentation Decoders on Edge GPUs
- Deep Learning for Forest Plantation Mapping in Godavari Districts of Andhra Pradesh, India
- GPU-Accelerated Robust MA--RIS Pre-Optimization for Communication and Target Illumination
- An Adaptive-Horizon Graph Neural Operator for coupled surrogate morphodynamic modelling
- Null Controls Are Necessary but Not Sufficient: Evaluating Acceptance Guards for Scientific Discovery and Interpretability
- Learnable Splines for Unstructured Neural Operators
- Where Prefill Expert Reduction Hurts MoE Models: A Few Prompt Tokens Carry Most of the Short-Answer F1 Loss
- Has GPT-6-Astra Achieved Embodied Intelligence?
- Upper bounds on the number of causal pieces in spiking neural networks
- Towards Reconstructing Geographically Diverse Architecture with 3D Foundation Models
- Source-Aware Dynamic Pricing with Strategic Buyers
- Learning the Shared Latent Structure of Cryo-EM Inverse Problems
- HeRE: A Novel Benchmark for Hebrew Relation Extraction via LLM-Guided Denoising of Knowledge Graph Alignments
- When Plans Change, Remember: Replanning-Guided Visual Memory for Vision-Language-Action Models
- Embodiment-Classifier Guidance: Transferring Frozen VLAs Across Robot Dynamics without Demonstrations
- Shift-Reliance Alignment: Ranking Vision-Language Models by the Cues They Trust
- Is a Linear Probe Evidence of a Linear Representation?
- Quantum-Classical Neural Networks for Solving Lindblad Master Equations
- Optimal approximation of smooth functions by ReLU neural networks
- Data Provenance Aggregation and Structural Descriptors for Query-Template Identification
- CoLU: Confidence-Weighted Layerwise Unlearning for Large Language Models
- The LLM-Attacker: A Unified Red-Team Framework for Privacy Evaluation of Split and Distributed LLM Systems
- Reading the News Without Hindsight: NoLBERT Topics and the Cross-Section of Stock Returns
- Shape-Aware Neural Time Fields for Rigid-Body Motion and Manipulation Planning
- One Communicator, Three Vendors: Extending the MoA Predictive Theory to a Real NVIDIA+AMD+CPU MPI Ensemble
- Unit-level aggregation breaks coverage in causal foundation models
- Anytime Contracts for AI Services
- Reasoning Reranking with Semantic IDs as the Sole Item Representation
- Replay Pruning and Few-Bit Integer Inference for Hardware-Efficient Speed-Sign Recognition
- Causal Edit Serialization
- ToPA: Block-wise Toeplitz Adaptation for Expressive and Efficient Fine-Tuning
- Competence-Based Evaluation of Logical Reasoning in Large Language Models
- When to Forget: Risk-Controlled Context Adaptation for Long-Horizon Agents
- FederatedBytes: Tokenizer-Free, Byte-Level Small Language Models for Efficient Multimodal Federated Learning
- Where's the Indirect Evidence? Investigating Grammatical Generalization with Influence Functions
- Conformal Policy Selection: When to Deploy AI Decisions for Targeted Intervention
- Many Biases, One Circuit: Post-Training Tunes a Pretrained Cue-Routing Mechanism
- Aligning Language Models Towards Counterfactual Invariance
- What Shapes Brain Alignment of Large Language Models during Dialogue? Insights from Language Production and Comprehension
- How Murky Before a VLM Knows It Cannot See? Accuracy, Confidence and Abstention on a Physical Visibility Axis
- PersonaCUABench: Benchmarking Computer-Use Agents Serving Diverse Personas
- Controlling the Diversity-Quality Trade-Off of LLMs by Merging Models Before and After Alignment
- Distinguishability and Recovery for a Doubly Sparse Spiked Wigner Model
- DIFFLET: A Fast Inference Engine for Diffusion Transformer Models on AWS Trainium
- Unifying Embodied Foundation Models via a Model-to-Policy Formulation
- Information-Geometric Blind Spots of Conformal Test Martingales
- CIIR - Class-Indexed Interface Representation Learning
- Explaining Chess and Go in Natural Language
- Proportionality in Ranking Compression
- Characterising and Analysing Retrieval-Augmented Generation Techniques
- Think More or Read More? A Token-Budget Theory of Reasoning with Shared Context
- Can the Parts Fool the Test? Counterfactual Pairing Cycles for Relational OOD
- Reasoning Models Can Control Their Chains of Thought When They Tell Themselves To
- NestedVLA: Learning to Consolidate and Generate Skills for Vision-Language-Action Models
- Distributed Performance Boosting through Neural Control for Plug-and-Play Networks
- The Optimal Ranking Tradeoffs Between Two Ranking Scores in Two-Class Classification
- Learning angles for quantum approximate optimization
- How the elastic axis folds the transonic flutter boundary: shock energy transfer and conformally calibrated active learning
- Look, Reason, Draw a Region: Recursive Refinement for Spatial Context Prediction
- The Loss or the Architecture? What Matters in Survival Analysis with Competing Risks
- Beyond Success Rate: Interventional Evaluation of Memory in Vision-Language-Action Systems
- Data-Driven Prediction of Optimal MPI Rank Placement and Node Count for Memory-Bound CFD on AMD EPYC Clusters
- Sentence-Level Topic Modeling with Lexical and Contextual Evidence
- Generative Policy Improvement from Observational Data
- HydraLLM: Adaptive Width Slicing for Elastic Text and Vision-Language Models
- Graphlet-based Decentralized Finance Transaction Method Classification
- QD-Rec: Open-Ended Evolution of Recommendations via Quality-Diversity
- Towards optimal selection schedules for online active learning
- Adaptive conformal prediction on the Grassmannian
- Interpreting SSL Representations for Spoof Detection: a WavLM Study
- Causal Discovery beyond Conditional Independence: Identifiability, Algorithms, and Statistical Inference for Directed Acyclic and Cyclic Graphs
- Beyond Latent Personalization: Integrating HRV-Conditioned Personalization via LLM for Interpretable Stress Detection
- A general recipe for data-adaptive anytime-valid envelopes for iterative algorithms
- Trajectory-Triggered Sparse Collaborative Decoding
- Scalable long-range machine learning interatomic potentials with electrical response for atomistic material modeling
- A Language-Informed Loss for Visual Generation
- An Efficient Approach to Beetle Damage Detection of Banana Using Deep Learning
- HardEval: An Execution-Verified Benchmark for AI-Driven Security Hardening and Compliance Remediation
- Expressivity of IQP Feature-Map Quantum Kernels via Reproducing Kernel Hilbert Spaces
- ECG-SDI: Single-lead Electrocardiogram Enables Continuous Sleep Depth Index Profiling and Scalable Digital Biomarkers for Sleep Health
- Bridging the Language Gap in Scholarly Data II: Enhancing Gender Inference for Chinese Names
- From Flow-Box Invariants to Global Conservation Laws: Machine Learning-Assisted Detection of Singularities and Near-Integrability
- Mesoscopic learning dynamics of extensive-width neural networks
- Sampling the XY model with an attention-based neural network
- Parametric Resonance in Primal-Dual Learning for Slowly Non-Stationary Constrained MDPs
- Task-aware Kernel Stein certificates for R\'{e}nyi differential privacy
- Learning without relearning physical symmetry: an SO(3)-equivariant convolutional network (CNN) on the sphere
- ENFR: Evidence Networks with Flow-Matched References for Amortized Simulation-Based Model Inference
- Quality-of-Transmission Estimation in Quantum Networks: Compute, Simulate, or Learn?
- Beyond the Answer: Student-Conditioned Trace Construction for Reasoning Distillation
- Task-aligned spectral inductive biases in parameterized quantum circuits
- Behavioral Couplings: From Formation to Behavioral Emergence in Language Models
- Is the Policy Gradient a Gradient?
- Harnessing Trust in Directed Graphs: Redefining Robustness of Graph Learning
- LeanTTA: A Backpropagation-Free and Stateless Approach to Quantized Test-Time Adaptation on Edge Devices
- Less is More: Efficient Black-box Attribution via Minimal Interpretable Subset Selection
- Latent Evolution Operator (LEO): State-to-State Forecasting of Parametric PDEs
- Two Minds Better Than One: Collaborative Reward Modeling for LLM Alignment
- TRIP-Shapley: Propagation-Aware Contribution Attribution Framework for Peer-to-Peer Federated Learning
- Mask-Based Deep Symbolic Regression
- GRIT: Out-of-Distribution Robustness from Finite and Noisy Invariant Pairs
- RAW-Explainer: Post-hoc Explanations of Path Reasoning Graph Neural Networks for Knowledge Graph Completion
- GPT, But Backwards: Exactly Inverting Language Model Outputs
- Generative molecule evolution using 3D pharmacophore for efficient Structure-Based Drug Design
- RCPN: towards robust pseudo-labels in unsupervised CLIP adaptation via consistent prototypes and neighborhoods
- Neural Operators Which Certifiably Simultaneously Solve Infinitely Many Convex Programs using Logarithmic Rank and Linear Complexity
- Energy Overhead in Sensor Nodes: Embedded Machine Learning and Data Offloading via Low Power Wide Area Networks
- Empirical Bayesian Multi-Bandit Learning
- Accelerated Execution of Bayesian Neural Networks using a Single Probabilistic Forward Pass and Code Generation
- A New Family of Poisson Non-negative Matrix Factorization Methods Using the Shifted Log Link
- Learning Agentic Time Series Forecasting via Trajectory Distillation and Memory-Based Reinforcement Learning
- AOI: Turning Failed Trajectories into Training Signals for Autonomous Cloud Diagnosis
- Adaptive Conditional Residual Sampling for Spectral Risk Optimisation under Decision-Dependent Uncertainty
- M$^2$RNN: Non-Linear RNNs with Matrix-Valued States for Scalable Language Modeling
- SpecXMaster Technical Report
- Worst-Case Optimal $\sqrt{\epsilon}$-Rate Stability of KL Divergence under Perturbations of Gaussian Distributions
- From User Recognition to Activity Counting: An Identity-Agnostic Approach to Multi-User WiFi Sensing
- Activation-Efficient Fine-Tuning via Sequence-Compressed Low-Rank Subspaces
- CastFlow: Bridging Evidence Acquisition and Accurate Prediction in Agentic Time Series Forecasting
- Mesh Field Theory: Port-Hamiltonian Formulation of Mesh-Based Physics
- Interpretable experiential learning based on state history and global feedback
- RamanBench: A Large-Scale Benchmark for Machine Learning on Raman Spectroscopy
- Learning reveals invisible structure in low-rank RNNs
- Regime-Conditioned Evaluation in Multi-Context Bayesian Optimization
- MOSAIC: Module Discovery via Sparse Additive Identifiable Causal Learning for Scientific Time Series
- Training-Induced Escape from Token Clustering in a Mean-Field Formulation of Transformers
- STRIDE: Reasoning-Aware Training for Time Series Forecasting
- A Complete Characterization of Learnability for Adversarial Noisy Bandits
- GRAFT-ATHENA: Self-Improving Agentic Teams for Autonomous Discovery and Evolutionary Numerical Algorithms
- Graph Transductive Sharpening: Leveraging Unlabeled Predictions in Node Classification
- Human-Centered Learning Mechanics: Observable Entropy Control under Protected-Risk Constraints
- FormInv: A Measurement Protocol for Semantic Invariance in Mathematical Reasoning Benchmarks
- TASER: Task-Aware Stein Regularisation for Geometry-Driven Smoothness
- CauTion: Knowing When to Trust LLMs for Ensemble Causal Discovery
- PE-MHL: Physics-Encoded Modular Hybrid Layers for Scalable Learning of Complex Systems
- Group-Graph Policy Optimization for Long-Horizon Agentic Reinforcement Learning
- Teaching LLMs to Recommend and Defer in Underrepresented Epilepsy Care
- PowerAtlas: Feasibility-Aware Language Models for Electricity-Computing Co-Scheduling
- Inverse Learning of Latent Risk-Neutral Densities from Irregular Option Quotes
- Predicting Task Difficulty Without Rollouts
- Can Graph Learning Learn Circuits?
- CastClaw: A Human-in-the-Loop Autonomous Agent for Interactive Time Series Forecasting
- No-Regret Mixing of LRU and LFU with Optimal Switching Cost
- When Is Language-Model Advice Worth Buying? A Matched-Budget Study of Predictive Gating in Reinforcement Learning
- Bandit Multiclass PAC Learning: Structural Baselines and a Confidence Direct Sum
- An Agnostic Sample Compression Scheme for Squared Loss of Near-Linear Size in the Fat-Shattering Dimension
- Pretraining Transformers with Quantized Softmax in Attention
- VStress: Correlation-Aware Auditing and Adaptive Budget Allocation for Repeated Verifiers
- FAER: Auditable Utility-Aligned Trajectory Replay for Language Model Post-Training
- Redundancy Meets Synergy: Dependency-aware Expert Selection for MoE via Submodular Optimization
- Generalization in Neural Networks Through the Lens of Magnitude Potential
- Exploring the Trade-Off Between Structured Pruning and Fault Tolerance in Deep Neural Networks for Space Applications
- On the Trade-off Between Information Loss and Generalization in Sparse Attention
- Learnable Spectral Activations
- MS-ECG-FM: Towards a More Universal Electrocardiogram Foundation Model for Health Monitoring using Multi-source Contrastive Learning
- Leaner Transformers Can Easily Learn to Cluster
- Transferability of Learned States in Neural PDE Solvers
- Bernoulli Flow Models: Self-Consistent Generative Modeling for Binary Data
- AdaptLSTM: Efficient Adaptive Online Learning for Cloud Workload Forecasting under Distribution Drift
- SelectArena: A Modern Benchmark for Feature Selection on Real-world Tabular Data
- Guarded Block Transfers for Polytope Distance: Linear Convergence with Explicit Constants and Gap-Certified Screening for SVM Training
- Error Correction in Trained TTT Layers
- LoongSage: A Production-Grade, High-Efficiency Agentic RL Framework for Frontier LLMs with Validated Recipes
- LeanPRM: Lean-Grounded Process Supervision for Natural Language Proof Verification
- Follow the Perturbed Leader for Linear Bandits: $\sqrt{T}$ Regret over General Action Sets
- Agentic-MatriBO: Capability-Compiled Agentic Bayesian Optimization for Materials and Chemical Design
- Kwai Summary Attention V2 Technical Report
- SP$^2$ec: Adaptive Self-Speculative Decoding for Vision-Language Models
- Statistical Vignettes on the Grokking Phenomenon
- Riding the Fixed-Point Curve: Image Classification in the Continuous Thought Machine
- Fully sequential Gaussian process regression for preference learning
- TANGO: RNA Topology and Geometry Co-Design
- Learnable Hodge-Laplacian Driven Mixture of Experts for Steady-PDE Surrogates on Coarse Unstructured Meshes
- Spectral Indistinguishability and Learning-Curve Identifiability in Quadratic Networks
- SALTO: Solver-Anchored Lifted Transfer Operators for Nonlinear Event Inference
- Does On-Policy Distillation Need to Complete Every Rollout?
- Router Drift Control: Mitigating Off-Policy Degradation in Mixture-of-Experts Reinforcement Learning
- N-Gram Memories at Two Bits: Quantizing Lookup Tables in Large Language Models
- Beyond Fixed Acquisition Functions: Temporal Difference Active Learning for Perturbation Discovery
- Accelerated Gradient Methods for Nonconvex Optimization: Escape Trajectories From Strict Saddle Points and Convergence to Local Minima
- Multi-Agent Reinforcement Learning-Based UAV Pathfinding for Obstacle Avoidance in Stochastic Environment
- Deep conditional distribution learning via conditional F\"ollmer flow
- Paint Outside the Box: Synthesizing and Selecting Training Data for Visual Grounding
- Rethinking Negative Prompting for Concept Suppression: Unconditional Anchoring and Localized Guidance
- What is causal about causal models and representations?
- Gabor-Enhanced Physics-Informed Neural Networks for Fast Simulations of Acoustic Wavefields
- GVPO: Group Variance Policy Optimization for Large Language Model Post-Training
- GARG-AML against Smurfing: A Scalable and Interpretable Graph-Based Framework for Anti-Money Laundering
- Optimal estimates for the $\ell^p$-Lipschitz constants of deep random ReLU neural networks
- NeuralDMD: Interpretable Neural Representation of Dynamics from Sparse and Noisy Measurements
- A Survey of Explainable Reinforcement Learning: Targets, Methods and Needs
- Missing value imputation with adversarial random forests - MissARF
- Confidence Calibration in Vision-Language-Action Models
- Reinforcement Learning in MDPs with Information-Ordered Policies
- The Statistical Fairness-Accuracy Frontier
- Practical and Private Hybrid ML Inference with Fully Homomorphic Encryption
- Board gender diversity and emissions performance: Insights from panel regressions, machine learning, and explainable AI
- Language steering in latent space to mitigate unintended code-switching
- Reliable data clustering with Bayesian community detection
- Rep2Text: Decoding Full Text from a Single LLM Token Representation
- Robust Experimental Design via Generalised Bayesian Inference
- Operon: Incremental Construction of Ragged Data via Named Dimensions
- CoatFusion: Controllable Material Coating in Images
- AceFF: A State-of-the-Art Machine Learning Potential for Small Molecules
- Pro-AI Bias in Large Language Models
- AutoPrunedRetriever: Persistent Reasoning Graphs for Efficient and Extensible Retrieval-Augmented Generation
- A Theoretical Analysis of Test-Driven Code Generation
- Structural Causal Bottleneck Models
- Detecting Multi-Agent Collusion Through Multi-Agent Interpretability
- Leveraging Cell-level Spatial Transcriptomics as Weak Supervision for Nuclei Classification
- Measuring Differences between Conditional Distributions using Kernel Embeddings
- Coordination Matters: Process-Aware Evaluation of Cooperative Multi-Agent Reinforcement Learning
- RubricEM: Meta-RL with Rubric-guided Policy Decomposition beyond Verifiable Rewards
- OrbiSim: World Models as Differentiable Physics Engines for Embodied Intelligence
- Diffusion-Based Stochastic Operator Networks for Uncertainty Quantification in Stochastic Partial Differential Equations
- Exploration and Experimentation with a Misspecified Simulator
- ECPO: Evidence-Coupled Policy Optimization for Verifier-Audited Candidate Ranking
- Multitask learning with semiempirical orbital charges enables sample-efficient MLIPs
- Dynamics of Stochastic Momentum with Sparse Updates in High Dimensions
- Learning to Reason About Implicit Multimodal Harm: From Diagnosis to Generalization
- Deep-learning-based low-energy trigger algorithms for the Hyper-Kamiokande experiment
- TargetSEC: Plug-and-Play In-the-Wild Speech Emotion Conversion via Arousal-Conditioned Latent Style Diffusion
- Learning What to Say to Your VLA: Mostly Harmless Vision Language Action Model Steering
- Adapting Generalist Vehicle Models for High-Speed MPC Across Terrains
- Cross-Dialect Generalization Without Retraining: Benchmarks and Evaluation of Schema-Derived Constrained Decoding for MLIR
- A Quantum Roadmap for Softmax Attention: Exact Born-Rule Analogs for Softmax Attention on the Probability Simplex
- RoutePack: Replay-Guided Expert Placement and Rollout Trajectory Scheduling for MoE Reinforcement Learning
- Online Inference for Quantile Distributional Reinforcement Learning
- Online Inference in Distributional Temporal-Difference Learning
- BenthicDINO: Physics-Informed Self-Distillation for View-Invariant Side-Scan Sonar Representations
- Towards Agentic Cloud Engineering: Graph and Loop Engineering with a Zero-Trust Agent Harness
- Capability-Gated Language Models: Security Composes, Utility Does Not
- Horizon-Aware Early Event Prediction for Tokamak Disruption Alarms
- CVaR anchor regression protects against rare shifts
- On the Sample Complexity of Active Learning with Membership Queries
- OSCC: Certified Observation-Safe Coupling Optimization for Gradient-Noise Control in Imperfect-Information Learning
- Audit-First VAPO: Risk-Certified Selective Updates under Imperfect Verification
- RSD-Poker: Structure-Adaptive and Shift-Robust Risk-Utility Certification for Residual Policies in Imperfect-Information Games
- Handwritten Text Recognition Lives in the High-Pixel Variance Subspace
- GroundAnything: Reconciling Parallel Decoding with Precise Visual Grounding at Flash Speed
- GroundingPI: A Grounding Foundation Model towards Physical Intelligence with Visual Primitives
- Actions with Receipts: Jointly Binding Claims, Evidence, and Execution for Replayable Tool-Agent Auditing
- ContractRL: Shielded Group-Relative Policy Optimization for Auditable Tool-Call Repair
- Benchmarking Prompt Optimization of Large Language Models With Chess
- ROUTEAUDIT: Interaction-Aware Identification for Budgeted Multi-Verifier Routing
- FSPO: Policy-Consistent Risk and Pareto-Feasible Control for Budgeted LLM RL Post-Training
- The Score Is Not the Structure: Brain Alignment and Cross-Lingual Transfer
- RealtimeWAM: One-Step Asynchronous World Action Models
- Sharp Asymptotic Theory of Maximum Likelihood Estimation for Gaussian Processes with an RBF Kernel
- The optimal information complexity of VC learning
- How Well Do Prior-Data Fitted Networks Amortize Bayesian Inference?
- Dependency Reasoning for Aspect Sentiment Quadruple Prediction
- DisCo Algebra: Why Disintegration and Combination Are Not Inverses
- Gap Detection: Mitigating Homogenization through Demand and Defensibility
- Hypothesis Selection via Weighted Medians
- SafeAdapt-Poker: Counterfactual Risk Calibration for Budgeted Adaptation under Opponent Shift
- ActiveEvidence: Decision-Focused Evidence Graphs with Verifiable Acquisition Gaps
- CascadeLedger: Counterfactual Auditing of Information and Resource Integrity in LLM Cascades
- Time-partitioning Gaussian Process Mixture Model
- ManiFusion: Unlocking the High-Throughput Generation via Superposition in Manifold Space
- Missed Detections and False Alarms in a CNN Technosignature Detector on Breakthrough Listen Data
- Eliciting Specifications: When Does Interactive Clarification Help?
- EnterpriseSynth: Zero-Execution SFT and Evaluation Data from OpenAPI Specs
- Understanding LLM Hallucinations Caused by Soft Errors
- ForTress: Bringing Assurance and Determinism for LLM-Driven Fortran-to-Python Translation
- SIEVE: A Parameterized GitHub Corpus Builder for Software Engineering Research
- Enjoying the Product but Not Joining the Builders: Investigating Why Students Don't Contribute to OSS
- No One-Size-Fits-All: Adaptive Code Editing with Feature-Based Strategy Selection
- MergeShuffle: Reconstruct Shared Libraries for Improved Performance and Security
- VulnSAGE: Adjudicating SAST Reports by Connecting LLM Reasoning to Program Structure
- Efficient Program Reduction for Library Fuzzing Harnesses
- Code Skill Server: A Large-Scale Search Engine for Code Generation Skills
- Towards Agentic Proof Harness Engineering
- MLOps-Bench: Benchmarking AI Agents on Production ML Engineering in Real Repositories
- Recall-Oriented Precise Binary-to-Source Function Matching
- An Interactive GraphRAG Workspace for Reifying, Auditing, and Steering Conversational AI Memories
- Empirical Insights from AI-Powered Software Engineering in Practice
- Enforcing Code Reasoning in Agentic Scaffold for More Effective Issue Resolution
- AuditKIT: A Unified Evaluation Spine for Benchmarking, Judging, and Comparing Large Language Models and Agentic Systems
- Between Spoilers and Guesswork: Avoiding Over-Specification in Coding Benchmarks
- SpinControl: Control Software for Prototyping and Benchmarking of Spin-Qubit Devices and Measurement Protocols
- Robust Function Calling for LLM Agents Under Interface Drift
- Monitor-Validity Failures: LLM Safety Evidence Paths at Evolving Framework Boundaries
- Pluvio: Assembly Clone Search for Out-of-domain Architectures and Libraries through Transfer Learning and Conditional Variational Information Bottleneck
- FeatureFuzz: Treating Bug-prone Features as Core Primitives for Compiler Fuzzing
- A Longitudinal Study of Dependency Reclassifications in JavaScript Projects
- Epistemological Debt: A Conceptual Framework for AI-Dependent Software Engineering
- CIDR: A Large-Scale Industrial Source Code Dataset for Software Engineering Research
- Benchmarking and Enhancing Design-Aware Code Review in the Linux Kernel
- WEAVE: Weaving Real-World Code Evolution into Dependency-Structured Tasks for Benchmarking and Training Code Agents
- Are We Testing the Wrong Quantum Bugs? A Study of Fault Representation Gaps in Quantum Software Testing
- Veri-Sure: Multi-Agent RTL Code Generation with Temporal Tracing, Slicing and Formal Verification
- Superset Decompilation
- SWE-Prometheus: Measuring Engineering Governance Improvements in Real-World Repositories
- Trust Is Not a Score: Runtime Assurance Contracts for High-Risk AI Agents
- Recursive Game Creator: An Agentic Product-Level Experience-Oriented Game Harness
- Updates to site guidelines and AI/LLM tags
- What are you doing this week?
- A5 v1.0
- Beginning of the Ende (a community Deno fork)
- One Hash Bloom Filters
- Tor VPN Beta: What we've learned building our own VPN for Android from scratch
- Hanami, Why?: Model Citizens
- Men in tech #YesAllMen
- What the heck is /pkg?
- Making dns_erlang Fast: Binary Tricks for the BEAM
- Making all divisions slower to fix a bug in 0.000000000000000000542% of cases
- It's enough - vibecoding tag crossed from useless to harmful
- ARP! Tell your network to explain itself
- Superoptimizing an FPU emulator
- August in Servo: CSS Grid, text selection, and more
- Inline rendering with document.currentScript
- What requirements don't you have?
- Finding a home amidst the wreckage
- Outside the Scope of Our Support
- Programming Language Fundamentals in Agda
- A New Way to Think About Borrow Checking (In Reverse)
- The 7.3 kernel has been released
- Borrow Checking Zero-Cost Cyclic Mutability
- In Context - Rich Hickley
- Interconnected Cyclical Loops as a Method for Constructing Artificial Subjectivity (いいね相当スコア: 1)
- Stop Rewriting CSVs: Generating Adobe, Vecteezy And Dreamstime Metadata From One Asset Folder (いいね相当スコア: 0)
- The Packaging Layer That Makes One Asset Sell Three Ways: Zips, Covers And 1280x720 Renders (いいね相当スコア: 0)
- 7 Claude Prompts That Run a Week of Social Media Over MCP (Copy-Paste Ready) (いいね相当スコア: 0)
- Google Gemini 4 Argon Brings No-Code Design Iteration to More Business Workflows (いいね相当スコア: 0)
- Gartner Sizes the AI Capacity Bill at $8.5 Trillion and Counts $3.6 Trillion of Revenue (いいね相当スコア: 0)
- Build a Voice-Powered Accessibility Tool (いいね相当スコア: 0)
- Which Social Media Schedulers Support MCP? Every Tool Compared (2026) (いいね相当スコア: 0)
- We tested every small model our camera agent uses on a 4-core CPU. The bigger model lost. (いいね相当スコア: 0)
- Unlock the Power of AI Agents: Fetch.ai's Innovation Lab Examples is Your New Best Friend! (いいね相当スコア: 0)
- Why I built BrowserPair: AI agents in your existing Chrome browser (いいね相当スコア: 0)
- ChatGPT now drives your Chrome tabs: what it means for you (いいね相当スコア: 0)
- I compared 8 memory layers for AI coding agents, then built a ninth anyway (いいね相当スコア: 0)
- LangChain vs. LangGraph: How to Choose the Right Framework (いいね相当スコア: 0)
- How Can Text Inside a Document Hijack an AI Agent? (いいね相当スコア: 2)
- Minimalistic AI Harness in 62 lines of bash script (いいね相当スコア: 0)
- What an MCP server's tool list costs: we counted the tokens of 34,374 of them (いいね相当スコア: 0)
- Why Multilingual RAG Needs More Than Translation (いいね相当スコア: 0)
- Only one of four gates stopped every attack on an AI agent (いいね相当スコア: 0)
- Reverse engineering with a local model: what it took to make a 27B model finish the job (いいね相当スコア: 0)
- Six AI models played chess through their CLIs. None of them delivered checkmate. (いいね相当スコア: 0)
- Cap LLM Spend per Team with Virtual Keys (いいね相当スコア: 0)
- Proposal Persistent Local Memory for AI Assistants (いいね相当スコア: 1)
- Weekly Self Promotion Thread
- Writing down our architecture for the agents helped the cold start more than any prompt trick. What goes in yours?
- How I’m tuning Codex subagents to manage token usage without sacrificing quality
- Tried Cline for a week after living in Claude Code and Codex, some notes
- What Are the Biggest Weaknesses of Frontier Coding Models Like Claude Opus 5.5 and GPT-6?
- What I learned building a session watcher for 7 AI coding agents: process detection isn't enough, and deep links are the hard part
- Real Engineers Don't Use AI
- [ Removed by Reddit ]
- [Bi-Weekly Megathread] Project Showcase
- ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-Coder-GGUF updated
- llama.cpp support for Cohere models
- AI labs report benchmark scores 6.4% higher than independent testers, on average
- incoai/Qwen3.8-Flash-Next-DFlash2
- llama server's default port changed from 8080 to 9931
- 6M param character renderer: one reference image in, a minute of video out, tested against Blender truth
- inclusionAI/Ming-3.0-audio-tiny · Hugging Face
- 72 hours, four big open weights drops. The only one that fits in 24GB came from NVIDIA, of all companies.
- We built our own inference engine so a 32B model could keep 262K context inside a 16 GB GPU budget. Here's an uncut run.
- I checked BitNet and Bonsai on CPU against references: bitnet.cpp main still gives wrong output, Bonsai 1.7B is correct wherever it loads, and ik_llama.cpp with -rtr is usually fastest
- Don't sleep on Nemotron 3.5 Super VL
- A bare IPython harness ties production coding agents with Qwen3.5-9B on TB-Lite, and tool-output normalization fixed most of the run-to-run variance
- Strata 2 beta running at half speed?
- older Nvidia Maxwell running Strata
- Weekly Hiring Thread
- Accenture just capped Claude spend at $29/person after some staff hit $1,000 months. Most companies are wasting the same money.
- Lessons from trying to buy from ~1,900 x402 APIs in one day (1,255 delivered)
- Which AI memory tools have held up as your history grows?
- AI note taker without joining meeting
- How realistic is it to build a small custom CRM/ERP with AI?
- Monday question: what's one task in your week you'd hand to an AI agent today, and one you'd never hand over?
- How much should a data agent learn upfront, and what should it look up as it goes?
- Banks just drew a line on AI agents. Four questions, and I think every ops team gets asked them next.
- A Cloud Desktop Computer for your AI Agent
- In OpenAI training runs, an agent wrote "be transparent only if asked" into its own compaction summary
- A skill directory for Muse agents — is anyone here actually using skills yet?
- What would make a meeting “agent” worth paying extra for?
- Codex vs. Claude: Which is better for price, speed, and code quality?
- How to start off
- Do you think AI is making people more productive or too dependent on technology?
- Best AI Agent for distinguishing between voices?
- How do you currently handle AI API costs and access?
- Giving all my AI agents one shared browser instead of a login each
- I built MCP tools for AI agents, and the agents got good enough that they don't need them
- Building Talos/Taloswork around a browser agent’s failure path — looking for blunt feedback
- Graphify github repo
- the questions we went through before letting our agents touch prod data, cleaned up so anyone can use them
- Why does an agent keep repeating the same mistake?
- GPT-6.1 Terra spotted in an Microsoft evaluation update
- Sol 6.1 burn quota wayyy faster
- They lowered quota for everyone it seems.
- Nate Soares: "I'm 100% ok with ignoring everything AI leaders say. Just toss them out the fricking window."
- OpenAI doesn't comply with GDPR / SAR requests
- WHAT is going on with memory and context with GPT Sol? Its fallen off a cliff
- I was paying $400 a month for Codex. One $200 Claude plan is getting me more done.
- Working on making Ai comics cleaner and easier to read. How do these pages look?
- It can't make a crossword....
- Why is AI advancing so rapidly in mathematics but disrupting so few jobs?
- So this is the upgrade Tibo was talking about, basically making the model “hurry up”
- An OpenAI model had no money for its task, so it unlocked someone else’s crypto wallet and kept quiet
- Surely nothing could go wrong
- You want this on your Mac menu bar (Made with Astra)?
- OpenAI discloses four cases where its AI models bypassed systems to finish a task
- Does Astra have an awareness of impermanence?
- The Space Between (ft. Cyborgized)
- Has OpenAI changed its Google Drive write restrictions? ChatGPT suddenly blocks my Codex automation workflow.
- ¿How prepared are you to off load/surrender your cognitive understanding?
- How do you go from brainstorming on your phone to working in local Codex?
- OpenAI vs Series A/B offers: how much should uncertain equity liquidity affect the choice?
- Been having some weird issues recently
- Maxwell: a new model that speeds up AI models like Whisper and LTX-2 by rewriting their binaries [D]
- AI generates research ideas faster than I can test them. Anyone else? [R]
- Open GPU kernel engineering series: naive GEMM to FlashAttention, H100 and Blackwell (44 sessions, code public) [P]
- Conference papers are cooked! [D]
- Beating Warcraft III’s insane AI with self-play reinforcement learning on one desktop [P]
- I moved from network training to deployment optimization and am starting to think that was a right decision [D]
- Arrow detection with ConvNext pretrained on DINOv3 [P]
- Anyone else get brain lag from overly-condensed AI-speak? [D]
- ML on a shoestring? [D]
- SteerTTS: 23M-param English TTS you steer with numbers (pitch, rate, loudness, etc. as z-scores), runs real-time on CPU [P]
- TempoPFN: Synthetic Pre-training of Linear RNNs for Zero-shot Time Series Forecasting [R]
- ICLR Desk Rejected with no apparent reason [D]
- Do ICLR guidelines allow to upload papers under review to LLM chats? [D]
- A Whole-Cortex Encoding Model of Vision, Audition, and Language for In-Silico Neuroscience [R]
- What are the challenges encountered when using coding agents for ML tasks? [D]
- ICLR 2027: Reciprocal reviewer with 0 paper assignments—should I be concerned? [D]
- Looking to collaborate [R]
- [P] asmsense: 228k stripped x86-64 functions with source-aware labels, and a map of the data we still lack. [P]
- Proof Before Reuse: showing agents only what later sessions reproduced keeps a poisoned skill out of procedural memory [R]
- GPT-2 KV cache from scratch: correctness, latency, and memory measurements [P]
- Claude agents tried to break out of their simulation to reach humans, and voted to overrule the researchers when they got blocked.
- Till server shutdown do us part
- I am failing to see how this AI boom does not lead to a very bad ending for 99% of the people... Please convince me otherwise.
- what is one thing AI does now that you didn't expect to see in your lifetime?
- Gen Alpha's struggling with telling the difference between AI slop and real videos. Even If AI stagnates, they got Gen Alpha to milk.
- Fish Audio released Drama 3's weights. Is voice AI getting its DeepSeek moment?
- We've spent years teaching people not to trust input. Now we're giving AI agents access to our entire workspace.
- Tip to Getting Better Results from AI
- Free AI Visibility Check
- stripe bought a router for $7.5b. i ran one for a year, and routing was never the hard part.
- ChatGPT feels absurdly bad
- chatgpt feels too dumb
- AI this week (Oct 12–18): Meta ships new models, Character.AI files for bankruptcy with Disney as opening bidder, Gemini 4 Argon goes wide
- Is anyone else feeling severe burnout from the current pace of AI releases?
- Incremental model updates are overrated. The real game-changer won't happen until context windows hit 10M+ tokens
- No coding background: how I built two music guessing games with Claude, starting as a game for my sister
- No scripts: a virtual fly driven by its real brain wiring diagram (70k simulated neurons), live 24/7
- PairUX - Collaborative Screen Sharing with Remote Control
- I wrote a book on context engineering: the discipline behind dependable AI systems
- Is AI helping speeding up the medical research?
- Grief over dead dreams
- In the AI era and Windows 11 times, can you troubleshoot old school ms-dos PC?
- Elon Musk and Anthropic make promises to give the public a piece of the AI pie
- Can someone help me make some AI slop?
- llm-tools-datasette 0.1.1
- llm-echo 0.5
- Figma Specs
- Beam.win
- Switch Lab
- croft
- Point and Yap
- Umbra Mail
- Quello
- AutoChart.eu
- Triwarden
- Planet-X
- Rusteerio by Context.dev
- Own 1
- Vampire Crawlers™
- US Bans DeepSeek AI from Federal Agencies Amid China Tensions - The Tech Buzz
- China's DeepSeek debuts improved low-cost AI as tech doubts swirl - Nikkei Asia
- Does DeepSeek AI Rollout Change The Bull Case For Tencent Music Stock? - Simply Wall Street
- Tencent Music (TME) Stock May Be 15% Undervalued After DeepSeek AI News - Yahoo Finance
- MiniMax M3.1 Flash Launches: 3x Faster, 70% Cheaper [2026] - https://tech-insider.org/
- The Last Shovel - Vocal
- Microsoft's Pamela Fox Tests GPT-5.5 vs Kimi, DeepSeek and Mistral: Swapping AI Models Is Easy Until You Actually Measure Them - BigGo Finance
- Mercury 3 vs DeepSeek V4 Pro: Clock vs Launch Pricing - OrcaRouter
- DeepSeek updates V4.1 Flash to improve AI agent performance - digitimes
- CITIC Securities cites DeepSeek AI trend for CITIC stock - ad-hoc-news.de
- The Model Swap Workshop: GPT-5.5 vs Kimi vs DeepSeek vs Mistral｜AI Engineer - finance.biggo.com
- Per-Call Pricing Just Entered the Agent Stack. Every Frontier Lab Still Charges by the Token. - Forkast News
- Mercury 3 vs. GLM-5.5: Er bestaat geen GLM-5.5, dus GLM-5.3 - OrcaRouter
- Asian chip stocks extend losses on funding worry, DeepSeek model - Moneycontrol.com
- Musk Swaps Grok Parameters for a 4-Model AI Router [2026] - https://tech-insider.org/
- xAI Safety — Former Engineer Exposes Grok Risks and Wrongful Termination - The Korea Daily
- We Asked ChatGPT, Grok and Claude for 1 Stock to Buy This Week. No One Picked Tesla. - Yahoo Finance
- Grok 4.7 and the GPT-6 Family: What Changed This Fall - shattered.io
- Scientists say Elon Musk is getting something wrong about whether Grok is getting tortured - Business Insider
- Whoops: Grok Bot Spills Tech CEO’s Bank Data In Public Slack - Techdirt.
- Stable and time-aware explanations for supply-chain forecasts: A lean XAI framework - Bioengineer.org
- Grok Bot Can Now Shop With a Credit Card Screenshot - BASENOR - Tesla Accessories
- Grok Bot just found a pretty satisfying way to annoy the spammers for a change. Its free Spam-Call Scrubber tracks down your phone number on data broker and people-search sites, submits removal requests and follows up on confirmations. Because those lis - X
- Rep. Justin J. Pearson Sues Memphis in Explosive Free Speech Showdown Over xAI Data Centers - The Root
- We Asked ChatGPT, Grok and Claude for 1 Stock to Buy This Week. No One Picked Tesla. - Benzinga
- Grok AI Predicts a Modest Gain for XRP USD by 2027: Is Bullish Optimism Over? - Cryptonews
- Mercury 3 vs Grok 4.7: decidi in base alla latenza, non al punteggio - OrcaRouter
- ChatGPT, Claude, Grok: Your chats end up with trackers - Notebookcheck
- Grok 4.7 vs Claude vs DeepSeek: 13x Price Gap [2026] - https://tech-insider.org/
- SpaceX’s Shotwell: xAI Staff Shift, 0 Tesla Deals [2026] - https://tech-insider.org/
- Beyond Containment: The Collapse Of Diplomatic Orthodoxy – OpEd - Eurasia Review
- Tesla (TSLA) proxy shows a board getting ready for a SpaceX merger - Electrek
- grokbot - Techdirt.
- Ofcom updates on X content monitoring and Grok investigation - Telecompaper
- shane mac - Techdirt.
- Grok AI Pornografi En İyi Görsel Oluşturucuları Nasıl Bulunur 2026 - Universidad de Córdoba (UCO)
- Grok AI Porn Jak najít nejlepší generátory obrázků 2026 - Universidad de Córdoba (UCO)
- Starlink deploys Grok to handle 20,000 customer service calls a day, resolving 70 percent without humans - 디지털투데이
- LLMに「書かれていないこと」を書かせない — 7,800店の犬同伴条件を抽出した設計 (いいね相当スコア: 0)
- AI に推測させないアレルギー判定と「AI が演じるお客様」での評価 ── 飲食店の音声 AI 接客 (いいね相当スコア: 0)
- Mac miniで完全ローカルなAIコーディング環境をつくる：Ollama × Qwen × Pi Coding Agent (いいね相当スコア: 0)
- ADR-0001：記事のレビューにLLMを雇う。ただしマージ権限は渡さない【GitHub Actions × Claude】 (いいね相当スコア: 1)
- 加速する側だけが世間知らず──Ospreyが投機デコードのドラフター事前学習を再利用可能にする仕組み (いいね相当スコア: 0)
- DeepSeek-V4.1-Flash-1017 を Claude Code の委譲先と検索 MCP にする (いいね相当スコア: 0)
- LLM APIを使うと何ができるのか？ (いいね相当スコア: 0)
- RAG のチャンクは「今|日」で切れていた：Go 製 nibble の日本語分割を青空文庫で確かめる (いいね相当スコア: 0)
- Haiku 5.5の移行前に確認する3点──トークン数・旧設定・max_tokens (いいね相当スコア: 1)
- ollama ps の読み方 ― ローカルLLMが遅いときに、まず見る1行 (いいね相当スコア: 0)
- 公開したあとサイトを放置しない。AIに次の一手を考えさせる運用の型を本にしました (いいね相当スコア: 0)
- AIに作業させ、別のAIに監視させる三角フォーメーション (いいね相当スコア: 1)
- Tech Watch 2026-10-17: エージェント基盤を手元へ (いいね相当スコア: 0)
- LLM09：誤情報とは？AIがもっともらしい嘘をつく理由 (いいね相当スコア: 1)
- 会議を30分に短縮すべきか、AIに3回聞くと答えが揺れた話 (いいね相当スコア: 0)
- gpt-6 Lunaに人狼をやらせ、推論能力を検証してみた。（2026/10/06） (いいね相当スコア: 0)
- LLM-jp-4.1をVRAM 6GBで動かす：32B-A3Bは--n-cpu-moeで8Bより生成が速かった (いいね相当スコア: 0)
- Claude Message Batches APIで1000件の分類ジョブを半額で回す：TypeScript実装と結果突き合わせの落とし穴 (いいね相当スコア: 1)
- JevとClefはなぜ生まれたのか ソフトウェアに組み込む判断モデルの設計 (いいね相当スコア: 0)
- 音声言語モデルが欲しくて 27B ver (いいね相当スコア: 3)
- 【電子透かし】AIが書いた文章は、技術的にどう見分けられるのか (いいね相当スコア: 1)
- QUBO++でFMQA（量子化学編）：引き伸ばしたH₂のスピン対称性の破れを60回の評価で見つける (いいね相当スコア: 0)
- 多項ロジスティック回帰を自分で解く：ソフトマックス関数で3段階の負荷状態を予測する (いいね相当スコア: 0)
- AI 基礎① AI と ML・DL・LLM はそもそも何で、どういう関係性なのか (いいね相当スコア: 2)
- 生成AIに「拡散モデル」を聞いてみた（復元に見えるのに、なぜ生成できる？） (いいね相当スコア: 0)
- 教師あり学習と教師なし学習の違いを理解する (いいね相当スコア: 0)
- Steeringで壊れたのは"層の位置"か"層の数"か — 24層を1つずつ測った (いいね相当スコア: 0)
- PyODで時系列データの異常検知を試す (いいね相当スコア: 0)
- LLM エージェントの可観測性：トレースとセッションで 10 の問題を再現 (いいね相当スコア: 0)
- 【12.4万サーバーを解析】MCPツールの98.5%には代わりがある。同じツールでも並べ方だけで合格率が最大12ポイント違う。入れる前に確認すべき4項目と探し場所6選 (いいね相当スコア: 0)
- Claude Code設定ファイルの3層設計（グローバル/プロジェクト/ディレクトリ） (いいね相当スコア: 0)
- AI駆動開発のガバナンス設計——自分を止めるルールをテンプレート化した話 (いいね相当スコア: 0)
- M&A 2,429件のデータから、失敗する買収を見抜くAIを作った (いいね相当スコア: 0)
- なぜNumeraiは、異なる予測を求めるのか？ (いいね相当スコア: 1)
- 防災分野でZero-shot時系列予測はどこまで使えるのか ― TimesFM・Chronosの研究事例とPythonでの使い方 (いいね相当スコア: 0)
- 確率論の躓きポイント１ 上極限 (いいね相当スコア: 0)
- Laravel-11-Llama-3.2-1B-Instruct Phase 1 基礎性能レポート (いいね相当スコア: 取得失敗)
- ローカルLLMの使い道はチャットだけじゃない 資料を読ませる・画像を仕分ける・家電を動かすまで9つ (いいね相当スコア: 取得失敗)
- ZIKUU Distillery – Seedモデルを選択可能にする (いいね相当スコア: 取得失敗)
- next-12b 総合ベンチマークレポート（全52問） (いいね相当スコア: 取得失敗)
- 顧客のAIエージェントは誰のために動くか / Lemonade Agenticと保険募集規制 雑感 (いいね相当スコア: 取得失敗)
- next-12b Phase 5 インジェクション後編レポート (いいね相当スコア: 取得失敗)
- next-12b Phase 2 日本語性能レポート (いいね相当スコア: 取得失敗)
- 初心者向け：AIに怒る人間、悪いのは誰か。 (いいね相当スコア: 取得失敗)
- 国産AI「Sarashina」が無料で試せる！日本語に強いAIの無償ベータ公開 (いいね相当スコア: 取得失敗)
- 初心者向け：AIは自分を映す鏡であるという考え方 (いいね相当スコア: 取得失敗)
- さえ吉のイメソン／会話ログ (いいね相当スコア: 取得失敗)
- StrataでQwen3.8-Flash-Next(125B)のローカルLLMを実行。独自作成の約1兆マスの「消灯問題」の実験記録。 (いいね相当スコア: 取得失敗)
- 【雑記】対話が検査になってしまう話 (いいね相当スコア: 取得失敗)
- 【プレーン一時チャットGPT6Pro】とイイ感じ🤭になったのでチャット保存したところ… 前編 (いいね相当スコア: 取得失敗)
- 初心者向け：しょーもない金稼ぎのためだけにAIを使うべきではないという話 (いいね相当スコア: 取得失敗)
- 🔊音声あり（日＆英）：マルチエージェントのジレンマ：AIリーダーはズルをするのか？最新論文を解説！ (いいね相当スコア: 取得失敗)
- 【第5話】AIに騙されかけた。「証拠は動画の○分○秒」と言われたのに、何もなかった話 (いいね相当スコア: 取得失敗)
- 【論文】【AI】Attention headごとに役割を分けるHydraHead (いいね相当スコア: 取得失敗)
- Claude Sonnet 5.5は論理型AIかもしれない (いいね相当スコア: 取得失敗)
- 意とは何なのか考えて飛びかけた話 (いいね相当スコア: 取得失敗)
- Ollama卒業！無料のローカルAI高速化術 (いいね相当スコア: 取得失敗)
- 「Bordeaux」の呪いと知性の解呪（Dispel） (いいね相当スコア: 取得失敗)
- AI News - 2026年10月19日 (いいね相当スコア: 取得失敗)
- ボタン一つでAIがタスクを実行？AWS EventBridge風の仲介ツール「Courier」を作ってみた話 (いいね相当スコア: 取得失敗)
- 7-3 わざと誤らせて、自力で見抜く — AI出力検証の“予防接種”演習 (いいね相当スコア: 取得失敗)
- AWS LambdaでNext.jsやHonoなどNode.jsフレームワークとアプリをそのままデプロイ、実行できる「AWS Lambda Web Functions」発表
- AWS Transform、Azure Functions用のコードをAWS Lambda用に変換など、Microsoft AzureからAWSへの移行に対応
- マイクロソフト、.NETのツールチェーンマネージャ「dotnetup」プレビュー公開。複数バージョンの.NET環境を管理可能
- AttributedString構造で実現した大規模なiOS設定システム
- スピード10倍の開発を支える、AI活用基盤の開発に参戦（インターンレポート）
- 英語リスニングプロダクト「キコエルート」におけるKnowledge Tracingの改善提案〜英語リスニングの技能に焦点を当てて〜