Best AI papers explained

506 Episodes

1. Direct Preference Optimization with Unobserved Preference Heterogeneity: The Necessity of Ternary Preferences

   Published: 24/10/2025
2. The Coverage Principle: How Pre-Training Enables Post-Training

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

   Published: 24/10/2025
4. Agent Learning via Early Experience

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

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

   Published: 22/10/2025
7. A Definition of AGI

   Published: 22/10/2025
8. Provably Learning from Language Feedback

   Published: 21/10/2025
9. In-Context Learning for Pure Exploration

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

    Published: 20/10/2025
11. Training LLM Agents to Empower Humans

    Published: 20/10/2025
12. Richard Sutton Declares LLMs a Dead End

    Published: 20/10/2025
13. Demystifying Reinforcement Learning in Agentic Reasoning

    Published: 19/10/2025
14. Emergent coordination in multi-agent language models

    Published: 19/10/2025
15. Learning-to-measure: in-context active feature acquisition

    Published: 19/10/2025
16. Andrej Karpathy's insights: AGI, Intelligence, and Evolution

    Published: 19/10/2025
17. Front-Loading Reasoning: The Synergy between Pretraining and Post-Training Data

    Published: 18/10/2025
18. Representation-Based Exploration for Language Models: From Test-Time to Post-Training

    Published: 18/10/2025
19. The attacker moves second: stronger adaptive attacks bypass defenses against LLM jail- Breaks and prompt injections

    Published: 18/10/2025
20. When can in-context learning generalize out of task distribution?

    Published: 16/10/2025