Daily Papers of 2026-09-24

  1. SpeakerMem-R1: Speaker-Centered Dual-Track Memory for Multi-Party Dialogue 100 upvotes, #1 of 2026-09-24
  2. The Past Frames the Future: Memory for Autoregressive Video Generation 64 upvotes, #2 of 2026-09-24
  3. Spatial-Interactor: Learning Spatial Reasoning through Interaction with the Observable Physical World 60 upvotes, #3 of 2026-09-24
  4. HappyWorld-Bench 56 upvotes, #4 of 2026-09-24
  5. Just-in-Time Memory: Learning to Curate Task-Adaptive Memory for LLM Agents 51 upvotes, #5 of 2026-09-24
  6. RewardVerse: Rubric-Guided Policy Optimization for Video Reward Modeling 40 upvotes, #6 of 2026-09-24
  7. Schrödinger's Code Repository: Have LLMs Learned SWE-bench or Memorized It? 31 upvotes, #7 of 2026-09-24
  8. PACT: From Credit Assignment to Critic Alignment 28 upvotes, #8 of 2026-09-24
  9. Hunyuan-A13B Technical Report 24 upvotes, #9 of 2026-09-24
  10. Capable yet Parsimonious: Extracting and Characterizing Hidden Chain-of-Thought in Frontier Models 22 upvotes, #10 of 2026-09-24
  11. Verifiable Hidden Dynamics Play: Generating Agentic RL Environments from Solved Mechanisms 22 upvotes, #10 of 2026-09-24
  12. All modalities are equal, but video is more equal: Closing the Cross-Attention Gap in Joint Video Generation 18 upvotes, #12 of 2026-09-24
  13. On the Diffusibility of High-Dimensional Latents 18 upvotes, #12 of 2026-09-24
  14. GeoPair: Geometry-Preserving Cross-Layer Factorization for Training-Free Transformer Compression 17 upvotes, #14 of 2026-09-24
  15. PackLab: A Comprehensive Framework for Developing, Training, and Evaluating MLLMs in Robotic Bin Packing 14 upvotes, #15 of 2026-09-24
  16. WhatWorkedBench: Benchmarking Experimental Understanding in AI Agents 14 upvotes, #15 of 2026-09-24
  17. InternW0: A Foundational Physical World Model for Efficient Real-World Interactions 14 upvotes, #15 of 2026-09-24
  18. MemBodied: Recurrent Associative Memory for Vision-Language-Action Models 14 upvotes, #15 of 2026-09-24
  19. MemoryAthena: Adaptive Routing over Latent and Generated Memories 11 upvotes, #19 of 2026-09-24
  20. StudentBench: AI and human tutoring yield equivalent GRE learning gains 10 upvotes, #20 of 2026-09-24
  21. Knowledge Pull Requests for Continual Document Authoring 8 upvotes, #21 of 2026-09-24
  22. FLEET: From Logits Entropy to Enhanced Trajectories in Text Generation 8 upvotes, #21 of 2026-09-24
  23. Self-Organizing Agent Teams Learn to Reason Together 6 upvotes, #23 of 2026-09-24
  24. Uranus: Building the Next-Generation Simulation Infrastructure for Embodied AI 6 upvotes, #23 of 2026-09-24
  25. Six Layers Less: Encoder Pruning for Whisper with Label-Free Recovery 6 upvotes, #23 of 2026-09-24
  26. The Linear Representation Hypothesis Needs a Group Action 5 upvotes, #26 of 2026-09-24
  27. Calibration as a First-Class Criterion in LLM Evaluation 4 upvotes, #27 of 2026-09-24
  28. EmbodiedSWE: Coding Agents for Long Horizon Dexterous Robotics 4 upvotes, #27 of 2026-09-24
  29. X-Planner: Event-Structured Task Planning for Embodied Intelligence 2 upvotes, #29 of 2026-09-24

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.