Best AI papers explained

512 Episoade

1. Self-improving LLM agents at Test-Time

   Publicat: 27.10.2025
2. KL-Regularized Reinforcement Learning is designed to Mode Collapse

   Publicat: 27.10.2025
3. How do LLMs use their depth?

   Publicat: 27.10.2025
4. Thought Communication in Multiagent Collaboration

   Publicat: 27.10.2025
5. Reasoning with Sampling: Base Models Outperform RL

   Publicat: 26.10.2025
6. Continual Learning via Sparse Memory Finetuning

   Publicat: 26.10.2025
7. Direct Preference Optimization with Unobserved Preference Heterogeneity: The Necessity of Ternary Preferences

   Publicat: 24.10.2025
8. The Coverage Principle: How Pre-Training Enables Post-Training

   Publicat: 24.10.2025
9. The Era of Real-World Human Interaction: RL from User Conversations

   Publicat: 24.10.2025
10. Agent Learning via Early Experience

    Publicat: 24.10.2025
11. Demystifying the Mechanisms Behind Emergent Exploration in Goal-conditioned RL

    Publicat: 22.10.2025
12. Rewriting History: A Recipe for Interventional Analyses to Study Data Effects on Model Behavior

    Publicat: 22.10.2025
13. A Definition of AGI

    Publicat: 22.10.2025
14. Provably Learning from Language Feedback

    Publicat: 21.10.2025
15. In-Context Learning for Pure Exploration

    Publicat: 21.10.2025
16. On the Role of Preference Variance in Preference Optimization

    Publicat: 20.10.2025
17. Training LLM Agents to Empower Humans

    Publicat: 20.10.2025
18. Richard Sutton Declares LLMs a Dead End

    Publicat: 20.10.2025
19. Demystifying Reinforcement Learning in Agentic Reasoning

    Publicat: 19.10.2025
20. Emergent coordination in multi-agent language models

    Publicat: 19.10.2025