Best AI papers explained

512 Épisodes

1. Self-improving LLM agents at Test-Time

   Publié: 27/10/2025
2. KL-Regularized Reinforcement Learning is designed to Mode Collapse

   Publié: 27/10/2025
3. How do LLMs use their depth?

   Publié: 27/10/2025
4. Thought Communication in Multiagent Collaboration

   Publié: 27/10/2025
5. Reasoning with Sampling: Base Models Outperform RL

   Publié: 26/10/2025
6. Continual Learning via Sparse Memory Finetuning

   Publié: 26/10/2025
7. Direct Preference Optimization with Unobserved Preference Heterogeneity: The Necessity of Ternary Preferences

   Publié: 24/10/2025
8. The Coverage Principle: How Pre-Training Enables Post-Training

   Publié: 24/10/2025
9. The Era of Real-World Human Interaction: RL from User Conversations

   Publié: 24/10/2025
10. Agent Learning via Early Experience

    Publié: 24/10/2025
11. Demystifying the Mechanisms Behind Emergent Exploration in Goal-conditioned RL

    Publié: 22/10/2025
12. Rewriting History: A Recipe for Interventional Analyses to Study Data Effects on Model Behavior

    Publié: 22/10/2025
13. A Definition of AGI

    Publié: 22/10/2025
14. Provably Learning from Language Feedback

    Publié: 21/10/2025
15. In-Context Learning for Pure Exploration

    Publié: 21/10/2025
16. On the Role of Preference Variance in Preference Optimization

    Publié: 20/10/2025
17. Training LLM Agents to Empower Humans

    Publié: 20/10/2025
18. Richard Sutton Declares LLMs a Dead End

    Publié: 20/10/2025
19. Demystifying Reinforcement Learning in Agentic Reasoning

    Publié: 19/10/2025
20. Emergent coordination in multi-agent language models

    Publié: 19/10/2025