Daily Papers of 2026-08-10

  1. SimWAM: A Simple World Action Model for End-to-End Autonomous Driving 105 upvotes, #1 of 2026-08-10
  2. SFT Conflicts, RL Coexists: A Theoretical and Empirical Analysis of Multi-Task Learning for LLMs 53 upvotes, #2 of 2026-08-10
  3. MatrAIx: Simulating the World with 8.3 Billion Persona Agents 44 upvotes, #3 of 2026-08-10
  4. Beyond Simply Environment Scaling: Designing Effective Environment Distributions for Multimodal Agent Learning 42 upvotes, #4 of 2026-08-10
  5. Small Foundation Models of Human Cognition and Behaviour 23 upvotes, #5 of 2026-08-10
  6. YOLO-PEFT: Parameter-Efficient Fine-Tuning on YOLO Family 20 upvotes, #6 of 2026-08-10
  7. When Activation Oracles Learn Not to Read: Concept-Specific Blind Spots in Fine-Tuned Oracles 18 upvotes, #7 of 2026-08-10
  8. DCAS: Decoupling CLI Agent Scaffolding to Internalize Planning across Scaffolds 18 upvotes, #7 of 2026-08-10
  9. StreamArena: Toward Continuous, Interactive, and Long-Horizon Agentic Streaming Video Understanding 16 upvotes, #9 of 2026-08-10
  10. Douyin Multimodal Embedding Model Technical Report 14 upvotes, #10 of 2026-08-10
  11. Efficient Knowledge Distillation for LLMs: Offline Top-K Logits and a Fused Chunked KL Loss 14 upvotes, #10 of 2026-08-10
  12. Addressable Memory for Video World Models 14 upvotes, #10 of 2026-08-10
  13. Reinforcement Learning with Evolving Rubrics as Rewards for Audio Reasoning 13 upvotes, #13 of 2026-08-10
  14. Relevant but Incomplete: Referential Dangling as a Paradigm-Level Failure Mode in Hard Prompt Compression 13 upvotes, #13 of 2026-08-10
  15. DuplexGen: Adaptive Synthesis of Human-AI Turn-Taking Dialogues 12 upvotes, #15 of 2026-08-10
  16. Uncertainty-Aware World Model for Aerial Image-Goal Navigation 12 upvotes, #15 of 2026-08-10
  17. Multi-Agent Forensic Reasoning for Generalizable Deepfake Video Detection 11 upvotes, #17 of 2026-08-10
  18. Round-Trip Consistency: Bidirectional Diffusion Models Can Predict Their Own Rollout Errors 10 upvotes, #18 of 2026-08-10
  19. Characterizing the Quality Profile of AI-Generated C++ in Production 10 upvotes, #18 of 2026-08-10
  20. The Optimizer Is the Agent: Reasoning-Driven Search across Prompts, Programs, and ML Workflows 9 upvotes, #20 of 2026-08-10
  21. Skaling: Chinchilla's Exponents Meet Kaplan's Coupling 9 upvotes, #20 of 2026-08-10
  22. Modular TTT: Rethinking Test-Time Training as Composable Modules 8 upvotes, #22 of 2026-08-10
  23. Enfold: Folding World Model Imagination into Predictive Representations for Ultra-Efficient Embodied Control 5 upvotes, #23 of 2026-08-10
  24. FATE: Frame-Level Audio-Visual Temporal Embedding 5 upvotes, #23 of 2026-08-10
  25. Adversarial Attacks for Good: A Survey of Proactive Protection across the Visual Content Lifecycle 5 upvotes, #23 of 2026-08-10
  26. OneEmo: A Unified Multimodal Reasoning Model for Emotion Perception, Understanding, and Interaction 5 upvotes, #23 of 2026-08-10
  27. PrivacyPeek: Auditing What LLM-Based Agents Acquire, Not Just What They Say 4 upvotes, #27 of 2026-08-10
  28. When Privileged Guidance Misaligns: State-Matched Routing and Contextualized Self-Distillation for Multi-Turn Agents 4 upvotes, #27 of 2026-08-10
  29. Do AI Personas Grow? Analyzing and Benchmarking Personality Evolution in LLM Agents After Life Events 4 upvotes, #27 of 2026-08-10
  30. Capek 0.5: An Execution-Centric Vision-Language Model for Embodied Intelligence 4 upvotes, #27 of 2026-08-10
  31. Zero Gap Is Not Restoration: Stratified Per-Question Probability Evaluation and Step-wise Mitigation of Benchmark Contamination 4 upvotes, #27 of 2026-08-10
  32. Towards Interpretable Foundation Models for Retinal Fundus Images 3 upvotes, #32 of 2026-08-10
  33. Complementary Matrix-Gated QKAN Fast-Weight Programmers for Quantum Dynamics Forecasting 3 upvotes, #32 of 2026-08-10
  34. Can MLLMs Decode the Creative Leap? Introducing C4 for Cross-Concept Understanding 3 upvotes, #32 of 2026-08-10
  35. CLIP-CC-Bench: Evaluating Paragraph-Level Video Descriptions in Video-Language Models 2 upvotes, #35 of 2026-08-10

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.