Daily Papers of 2026-11-18

  1. TeleChat to Xing4: Advancing Frontier Agentic Models on Domestic Chips and Frameworks 154 upvotes, #1 of 2026-11-18
  2. AscendGYM: A Cost-Aware Kernel Optimization Benchmark for Real-World Workloads on Ascend NPUs 91 upvotes, #2 of 2026-11-18
  3. Weight-Space Programming Doesn't Need Transformers 51 upvotes, #3 of 2026-11-18
  4. Multi-Branch Transformer 45 upvotes, #4 of 2026-11-18
  5. ForceHarness: Turning Force into Action and Physics for Frozen Frontier Models 40 upvotes, #5 of 2026-11-18
  6. Breath-Dx: A Time Series Classification Benchmark for Disease Screening with eNose 32 upvotes, #6 of 2026-11-18
  7. Self-Improvement Memory Enables Continual Recursive Self-Improvement 31 upvotes, #7 of 2026-11-18
  8. FastFrames: Video Generation at Interactive Latency 24 upvotes, #8 of 2026-11-18
  9. TCP4D: A Unified Tool-Center-Point Tracking and Generation Model for Robot Manipulation 23 upvotes, #9 of 2026-11-18
  10. Heave-Ho: Decentralized Long-Horizon Multi-Humanoid Collaboration 21 upvotes, #10 of 2026-11-18
  11. VideoNexus: Video Perception Beyond the Reasoning Context 20 upvotes, #11 of 2026-11-18
  12. Regularized Score Centering 13 upvotes, #12 of 2026-11-18
  13. π^4: A Foundation Model for 4D Scene Geometry and Motion 13 upvotes, #12 of 2026-11-18
  14. From Repositories to Training Environments: Scaling Long-Horizon Coding Agents with Environment Evolution Loop 12 upvotes, #14 of 2026-11-18
  15. Beyond Answering: Learning to Speak First and Act in Streaming Video 10 upvotes, #15 of 2026-11-18
  16. MedJev: A System One Vision-Language Decision Model for Fast and Effective Medical Image Understanding 10 upvotes, #15 of 2026-11-18
  17. FMT: Lossless Image Compression in a Fixed Number of Parallel Steps 10 upvotes, #15 of 2026-11-18
  18. MPEG-Diffusion: Mixed-Resolution Video Generation 8 upvotes, #18 of 2026-11-18
  19. CADMAGEDDON: High-Fidelity CAD Reconstruction with Local Inference 8 upvotes, #18 of 2026-11-18
  20. Agentic On-Policy Distillation Implicitly Reshapes LLMs' ReAct Patterns 7 upvotes, #20 of 2026-11-18
  21. VisualArtifactSpec 1.0: Evaluating Visual Artifacts under Task Specifications 6 upvotes, #21 of 2026-11-18
  22. Do LLM Agents Still Need to Be Told How to Self-Improve? 5 upvotes, #22 of 2026-11-18
  23. Glyph: An Interpretability Engine for AI-Driven Scientific Discovery in Clinical Data 4 upvotes, #23 of 2026-11-18
  24. The Noise Floor of a Quantization Comparison: Engine, restart, batch and kernel drift, measured against 8-bit and 4-bit effects in two decision models 4 upvotes, #23 of 2026-11-18
  25. ResWAM: Learning What Changes, Not What Stays, for World-Action Models 4 upvotes, #23 of 2026-11-18
  26. DNeo: Your Parallel Draft Model Needs Better Target Feature Encoding 4 upvotes, #23 of 2026-11-18
  27. HubEgo-100K: In-the-Wild Egocentric Video of Skilled Work for Robot Learning 4 upvotes, #23 of 2026-11-18
  28. AuXSpec: Learning to Autoformalize Program Specifications via Executable Postconditions 3 upvotes, #28 of 2026-11-18
  29. MillBill: A Machining Agent with Geometric Feedback 3 upvotes, #28 of 2026-11-18
  30. Attention at Absolute Zero 2 upvotes, #30 of 2026-11-18
  31. PanDWM: Panoramic 4D Driving World Models 2 upvotes, #30 of 2026-11-18
  32. The novelty exposed: ablation, reproducibility and transparency in 40 technical reports of open-weight large language models (LLMs) 2 upvotes, #30 of 2026-11-18
  33. TPHC: Improving Hyper-Connections with Tensor-Parallel Grouped Updates 2 upvotes, #30 of 2026-11-18
  34. M3OPD-Reasoner: An Open Recipe for Post-Training Multimodal Reasoning with Multi-Teacher On-Policy Distillation 2 upvotes, #30 of 2026-11-18
  35. WeightBench: Evaluating Language Models as Weight-Space Programmers 1 upvotes, #35 of 2026-11-18
  36. Unreasonable Confidence from Simulator Misspecification: Directional Robustness on the SPEED+ Sim-to-Real Benchmark 1 upvotes, #35 of 2026-11-18
  37. MInfoVisQA: A Multilingual Benchmark for Information-Rich Visual Question Answering 1 upvotes, #35 of 2026-11-18
  38. GameRealm: Grounding Visual Learning in Interactive Games 1 upvotes, #35 of 2026-11-18
  39. CONGA: Compiling Gaussian Fiber Geometry for Structured Bayesian Inference 0 upvotes, #39 of 2026-11-18

Data: hysts-bot-data/daily-papers-stats and the Daily Papers API. Open data: tardellirs/paper-pulse-data. Sister project: Model Pulse, the download history of every model on the Hub.