Archive
Everything we've scored that's past the 48-hour window — open to everyone, no account needed. The live feed only shows what's moving right now; this is the record behind it.
- 19Verifiable by Construction: Claim-Level Evaluation of Verbatim Citation in Clinical Question Answering
- 19A Response Theory Probe for Learned Stochastic AI Simulators, Tested on Lorenz-63
- 19MC-Sparse: Deconstructing and Closing the Dense-Sparse Attention Gap in Diffusion Transformers
- 19Discord-Nitro-Generator — 🚀 Advanced Discord Nitro Generator and Checker (2026). Fast, multi-threaded, and proxy-supported free Discord Nit
- 19Mind2Dialogue: Training Human-Aware Language Models by Simulating User Mental States
- 19Discovery Foundation Models: Toward Open-Ended Discovery Intelligence
- 19Disentangling Representation Evolution in Transformers through Directional Decomposition
- 19Sharpen Without Search: On-Policy Distillation of Sequence-Level Power Distribution
- 19H-JEPA: End-to-End Learning of Hierarchical World Models for Visual Planning
- 19Block Disentanglement in CRL: Bridging Identifiability and Visual State Estimation
- 19Finding Gaussian Structure in Bosonic States
- 19A Chosen Future Can Still Be Rewritten: Causal Writability in Video Models
- 19The Router Within: Eliciting Native Skill Routing from a Frozen LLM
- 19TAPDreamer: Transferable Adversarial Patches for World Action Models
- 19Stellar Colosseum: A Many-Agent Harness for Long-Horizon Research in Mathematics and Theoretical Computer Science
- 19EIP Editing Office Hour (EIP + ERC ) Meeting #113, September 29, 2026
- 19Paradee: Distilling Kokoro-82M into an 8M-Parameter Single-Voice Text-to-Speech Model
- 19Bellman Policy Optimization
- 19Private online learning and prediction for Littlestone classes
- 19Deep Learning for Sleep Heart Rate Estimation from Accelerometers: Toward Population-Scale Cardiac Insight Without Optical Sensors
- 19TasteVal: Measuring the Experimental Research Taste of AI Systems Against Human Experts
- 19PlotGround: Grounding Plot Digitization in Real Scientific Figures and Their Source Data
- 19CLIFT: Conformal Self-Verification for Web Agent Training and Test-Time Scaling
- 19MemPilot: Orchestrating On-Demand Multimodal Memory Curation for LLM Agents
- 19UniSlider: Perceptually Uniform Sliders for Continuous Image Editing
- 19Towards Looped Models Done Right, Part II: Rethinking at Fixed Points
- 19Direct Intermediate Initialization for Tilted Diffusion Samplers
- 19Your .mcp.json probably has a live API key in it
- 19Recursive Video In-Context Learning for Agentic Robot
- 19Learning to Read the Contextual Tokens in Diffusion Transformers
- 19BiasFlow: Geometric Monitoring and Backbone Regularization for Spurious Feature Reliance
- 19Base Models Can Reason By Taking a Cue From Training Data
- 19One Figure, Every Canvas: Editable Flowchart Relayout via Agentic Pipeline
- 19EIP Editing Office Hour (EIP + ERC ) Meeting #115, October 07, 2026
- 19Gleam has mirrored its source code on tangled (an AT-protocol based forge)
- 19Are we the abstraction? AI and the future of software engineering
- 19Type Diversity Enables Transformers to Generalise Compositionally
- 19SAS: Simple Attention Sparsification via End-to-End Optimization of Context Ranking
- 19Rethinking Heterogeneous System Disaggregation for Subquadratic Attention
- 19Calling the AI bluff: Adding "Do not guess" cut made-up claims from 71% to 20%
- 19117 Ghost Errors: Anatomy of a Flaky AI Agent
- 19A Hybrid LSTM-XGBoost Framework for Multi-Horizon Stock Return Prediction Across Diversified Equity Portfolios
- 19CMA-OT: Hierarchical Expert Supervision for Dance-to-Music Generation
- 19Continue, Adapt, or Yield: In-Turn Adaptation to Overlapping Speech in Full-Duplex Agents
- 19Learning to Stop without Learning to Stop: Self-Supervised Confidence Training Improves Reasoning Efficiency
- 19Gap-free Differentially Private PCA for Gaussian Data
- 19Diffusion Models and Concept Formation
- 19First-Order Stationarity of Reverse Diffusions
- 19Kraken: LLM-based Speech-to-Speech Translation via Low-bitrate VQ and Dual-path Source Conditioning
- 19Statistical attribute alignment for black-box generative AI via output post-processing
- 19Robust Policy Optimization via Adversarial Importance Sampling
- 19Unified CT and MRI Pancreas Segmentation for Label-Efficient Cross-Modality Subregion Transfer
- 19DynSHAP: Towards Explainable Dynamic Survival Analysis
- 19User Model Extraction via Belief Self-Distillation
- 19New LoRA Skills Should Read but Never Write
- 19Quantile-based Loss Filtering for Outlier-Robust Stochastic Gradient Descent
- 19Transfer Learning for Evolving Domains
- 19RPC Standards # 34, September 7th, 2026
- 19A Ranking Approach for Measuring Calibration
- 19Groupoid-Based Internal State Representations for Reinforcement Learning with Local Symmetries
- 19Attention Quantization for Tabular Foundation Models
- 19Common-Mode Collapse and Recovery in Direct Feedback Alignment
- 19No Atlantic hurricanes by Sept. 12 breaks a 60-year record
- 19Compact Documentation for Coding Agents: A Benchmark, an Optimizer, and Why It Does Not Transfer
- 19Trust Guided Decision Transformer
- 19Label-Guided Knowledge Distillation for 3D-CNNs in Action Recognition
- 19Does the “Reflection” category in dotTrace show the time spent executing reflection in my own code?
- 19Strategically Diverse Sampling for Self-Training
- 19OC-GS: Gaussian Splatting for Irregular Turntable Capture
- 19LIBERO-PeRM: Benchmarking Personalized Robotic Manipulation
- 19Uncertainty and Explainability in Deep Rough Volatility: A Neural Information-Theoretic Posterior Approach
- 19Discovering state-of-the-art quantum circuits with large language models
- 19T2-z: Test-Time Optimization of Environment Latents for Robotic World Models
- 19CoDA: Co-Evolution of Data Engines and Action Policies
- 19Adapting for AI: How elementary teachers adjust their practices for an AI-integrated curriculum
- 19PathFM: Fast Generative Forecasting of Multivariate Trajectories
- 19DeepEdu-v1: Efficient and Scalable Agentic LLMs for Vietnamese Education
- 19OmniVerifier-C1: Decomposed Graded Verifier for Rule-Governed Generative World Modeling
- 19Rethinking Routing Exploration in Modern Mixture-of-Experts
- 19ASTRIL-MPC: Autonomous Traversal Framework of Articulated Tracked Robots with Language-Guided Neural-Kinematic MPC
- 19MLExpert: Towards a Generalist Agentic Model for Autonomous Machine Learning Engineering
- 19GeoState: Learning View-Decoupled Geometry States for 3D World Modeling
- 193DCodeVerse: Exploring 3D Worlds as Code
- 19Dexterity from Supervised Play
- 19Embodied-BenchForge: A Closed-Loop Agentic Workflow for Embodied Benchmark Construction
- 19Coding Agents with an Obstacle-Aware Harness for Safe Robot Manipulation
- 19Embedding Models Measure in Peculiar Ways
- 19Workspace Models: Lightweight Robotic Memory via Saliency-Driven Supervision
- 19Lean Graph Search for Formal Theorem Proving
- 19QGPINNs: A Physics-Informed Neural Network Framework for Nonlocal Differential Equations on Quantum Graphs
- 19FAMOS: Feed-Forward 3D Articulation Modeling from Sparse Observations
- 19Paint-Anything: Unified Any-Color Control for Image Generation and Editing
- 19ERCPMP-Gx: Endoscopic Image and Video Dataset for Morphological, Histopathological, and Genomic Characterization of Colorectal Polyposis
- 19How Does Distribution Shift Shape Pretraining Gains in Neural PDE Surrogates?
- 19Weight Pair Encoding: Inducing a Smaller Grammar in Neural Network Weights
- 19Quantifying Overclaiming Propensity in Frontier LLM Agents
- 19Multi-agent Scaling Across Disjunctive and Compensatory Tasks
- 19Robust Reinforcement Learning for Dynamic Treatment Regimes under Model and Measurement Uncertainty
- 19Unifying Models of Intergroup Hostility in Online Discourse
- 19Score Centering Stabilizes Off-policy Reinforcement Learning
- 19An Empirical Study of Harness Design for Coding Agents
- 19Evaluation Awareness Beyond English: Do LLMs Act Differently When Tested in Indic Languages?
- 19KnowGPO: Knowledge-Grouped Policy Optimization for Search Agents
- 19Designing Transformers for Continuous Diffusion Language Models
- 19ACEdit: Stable Multi-Turn Image Editing via Asymmetric Conditioning
- 19Generalization behavior of OPTQ and the role of regularization
- 19Twice Upon a Time: Complementary Representations for Time-Series Reasoning
- 19Online Learning via Learned Latent Bayesian Tracking
- 19Training Reasoning Models to Generate Diverse Scientific Ideas
- 19Learning Self-Orchestration in Decentralized Agent Swarms
- 19AVC-Reward: Event-Centric Audio-Video Consistency Evaluation
- 19MP-Bench: Evaluating Voice Agents as a Multiparty Conversation Participant
- 19JEPA-Anything: Learning Predictive Models across Different Worlds
- 19A team of Gemini 4 Argon agents beats Minecraft
- 19Learning to Steer LLM Agents and Multi-Agent Systems
- 19MindEval2: Clinically Grounded Patients and Rubrics for Evaluating Language Models in Mental Health Support
- 19Grid-Bench: Can LLMs Predict How Discrete Worlds Change?
- 19PosteriorBench: From Point Estimates to Posterior Matching in Evaluating Generative Inverse Solvers
- 19Aero Hand Open: A Simulation-Ready Tendon-Driven Hand for Dexterous Manipulation Learning
- 19Autonomous Research for Open-Ended Problems: A Case Study on Telecom Ticket Retrieval
- 19MAxBench: A Multinomial Concept Recovery Benchmark
- 19MexHat: A Dataset for Hate Speech Detection in Mexican Spanish Videos
- 19Learning a Size-Weight Frontier for Synthetic-Augmented Inference
- 19Knowledge That Is Linear Before Fine-Tuning Generalizes After It
- 19RetireOPD: Self-Retiring On-Policy Distillation for Agentic Reinforcement Learning
- 19Involving before Evolving: A Vision for Trustworthy Enterprise Digital Twin Engineering
- 19AI Consilium: a Benchmark of LLM Clinical Multidisciplinary Second Opinions
- 19Corrupt Plans, Clean Traces: Evading Chain-of-Thought Monitoring with Plan Injection
- 19Private Retrieval Augmented Generation via Random Projection
- 19Backward Compatibility in Zero-Bit Neural Audio Watermarking
- 19EAServe: Encode-Aware Disaggregated Serving for Multimodal Large Language Models
- 19UNWATCH: finding what a second tutorial actually adds
- 19Harm Laundering in GPT Models: Evidence That Gender Discrimination Is Transformed Rather Than Reduced Across Safety-Trained Generations
- 19GeoAAC: Geometry-Based Adaptive Action Chunking from Denoising Trajectories in VLA Policies
- 19Two Conformal Constructions for Adaptive Within-Document AI-Text Screening
- 19BeatGraph: Self-Supervised Heartbeat Graphs for Infant ECG Representations from the Home Environment
- 19Semantic Action Graph: A Shared Representation for Agent Grounding and Human Interpretation of Sports Highlights
- 19Anchoring Clinical Events in Time: UID-Preserving Multimodal Reconstruction and Source-Grounded Adjudication
- 19A Flow Matching Framework for Neural Representational Dissimilarity
- 19Calibrated RF-Fingerprinting Under Interference With Heterogeneous Transmission Protocols
- 19CanvasAnneal: Curriculum Reinforcement Learning for Diffusion Language Models
- 19On two proofs of $d^2$ mixing of weighted Dikin walks
- 19Agile-WAM: An Agile Tactile World Action Model for Contact-Rich Robot Control
- 19Expert-Space Exploration in MoE Reinforcement Learning
- 19Prediction-Powered Smoothing and Validation for Disaggregated AI Evaluation
- 19OPTED: On-Policy Fine-Tuning for End-to-End Driving using a Render-Free Teacher
- 19Learning between the peaks: sharp asymptotics for kernel ridge regression under power-law anisotropy
- 19RAFT: A Stateful Retrieval-Augmented Framework for Troubleshooting Agents
- 19Benign Loss Landscapes Can Coexist with Worst-Case Hardness
- 19Can You Check That? The Checkability Boundary for Local LLM Network Automation
- 19NEXT: Physics-Informed Neuro-Spectral Exponential Time Differencing Architectures
- 19Large Language Models as Falsifiers for Cyber-Physical Systems
- 19dQwen3.5: Hybrid-Attention Diffusion Language Models
- 19The git config Defaults Worth Overriding
- 19A Formal Limitation on Learning Human Language From Textual Corpora
- 19I analyzed 14,905 love letters without reading them
- 19HySTAR: Anchored Hypergraphs for Stable Credit Assignment in Cooperative Multi-Agent Reinforcement Learning
- 19MILER: Semantic Mid-Level Representation for Sim-to-Real Reinforcement Learning in Unstructured Autonomous Driving
- 19Designer-RSI: Evolving Procedural Memory from User Traffic for Agentic Graphic Design
- 19Blog: Survey of Optimizers
- 19Cross-sector generalization of accident-process role classification in occupational accident narratives
- 19Dynin-Robotics: Omnimodal Unified Diffusion Vision-Language-Action Model
- 19Video DeltaNet: A Video-Native Hybrid Attention for Livestream Video Generation
- 19“We have information that Moonshot distilled Fable for the development of K3”
- 19One Basis to Animate Them All: Gaussian Blendshape Distillation for Real-Time Avatars
- 19KaliBench: A Fine-Grained Benchmark for Cybersecurity Tool Use on Kali Linux with Runtime-Free Verifiable Rewards
- 19Reconstruct, Practice, Go Real: Guided Self-Improvement for Embodied Agents
- 19Embedding Prediction Helps Image Generation
- 19ScholarCatalyst: A Benchmark for Retrieving Papers That Inspire New Research
- 19SILSA: Sliding-Window Slice Latents for Topology-Preserving High-Resolution 3D Generation
- 19VISTA: A Visual Harness for Reasoning in an Interactive World
- 19TACO: Ternary Absolute-max Column-wise One-sparse Optimizer for LLM Fine-Tuning
- 19FERPO: Forward Entropy-Regularized Policy Optimization
- 19Hierarchical Continuous Diffusion Language Models
- 19Cost-augmented Schrödinger bridges on graphs are exactly solvable: a Feynman-Kac tilt replaces learned control
- 19ConceptGuard: Benchmarking Context-Sensitive Unlearning in Large Language Models
- 19Information on trajectories: martingales and random times
- 19The Missing Primitive: Diagnosing and Repairing Mathematical Reasoning in Large Language Models
- 19Trust the Direction, Search the Step: Zero-and-First-Order Methods for LLM Fine-Tuning
- 19G-CARL: Grounded Checklist-Aligned Reward Learning for Patient-Oriented Medical Report Interpretation
- 19Generative modeling of intrinsically disordered protein regions by reinforcing sparse autoencoder features
- 19Compile by Training: Turning Natural-Language Specifications into Local Neural Functions
- 19Clean Engineering, Unstable Measurement: A Preregistered Reliability Failure of Black-Box LLM Observers on Shared Endpoints
- 19ESPO: Error-Structured Prompt Optimization via Diagnose, Diversify, and Stabilize
- 19DMAD: Distribution Matching as Adversarial Distillation for Fast Visual Generation
- 19CodeMidas: Scaling Agentic Coding RL Environments from Code Itself
- 19Value-Sensitive Delegation in Everyday AI Agent Use: Evidence from OpenClaw
- 19AutoDesign: Meta-Harness Optimization for Long-Horizon Agentic Design
- 19Higher-Order Molecular Grammars for Generative and Foundation Models in Chemistry
- 19OmniScientist: An Omni-Modal Omni-Discipline AI Scientist
- 19Legibility is Not Interpretability: Comparing Judged and Actual Importance in Chain-Of-Thought Reasoning
- 19Logos: An Agent Harness on a Cross-Process Bus
- 19Decoding Looped Transformers Better for (Almost) Free
- 19A Unified and Constrained View of Regularization-Based Robust Reinforcement Learning
- 19One Editor, Many Edits: A Unified Training-Free Framework for Diverse Video Editing
- 19HumanTracker: Towards Comprehensive and Human-Aligned Motion Tracking Benchmark
- 19Robust PAC Learning of Concurrent Stochastic Games
- 19$TCP_α$: Margin-Controlled Confidence estimation for reliable Music Information Retrieval
- 19Defensive Boosting for Online Probabilistic Forecasting
- 19SoftServe: A Scalable Quasi-Newton Method for Deep Learning