Best AI papers explained

440 Episodes

1. Training a Generally Curious Agent

   Published: 6/12/2025
2. Estimation of Treatment Effects Under Nonstationarity via Truncated Difference-in-Q’s

   Published: 6/12/2025
3. Strategy Coopetition Explains the Emergence and Transience of In-Context Learning

   Published: 6/12/2025
4. Emergent Misalignment: Narrow finetuning can produce broadly misaligned LLMs

   Published: 6/11/2025
5. Agentic Supernet for Multi-agent Architecture Search

   Published: 6/11/2025
6. Sample Complexity and Representation Ability of Test-time Scaling Paradigms

   Published: 6/11/2025
7. Aligning with Human Judgement: The Role of Pairwise Preference in Large Language Model Evaluators

   Published: 6/10/2025
8. LLMs Get Lost In Multi-Turn Conversation

   Published: 6/9/2025
9. PromptPex: Automatic Test Generation for Prompts

   Published: 6/8/2025
10. General Agents Need World Models

    Published: 6/8/2025
11. The Illusion of Thinking: Understanding the Strengths and Limitations of Reasoning Models

    Published: 6/7/2025
12. Decisions With Algorithms

    Published: 6/7/2025
13. Adapting, fast and slow: Causal Approach to Few-Shot Sequence Learning

    Published: 6/6/2025
14. Conformal Arbitrage for LLM Objective Balancing

    Published: 6/6/2025
15. Simulation-Based Inference for Adaptive Experiments

    Published: 6/6/2025
16. Agents as Tool-Use Decision-Makers

    Published: 6/6/2025
17. Quantitative Judges for Large Language Models

    Published: 6/6/2025
18. Self-Challenging Language Model Agents

    Published: 6/6/2025
19. Learning to Explore: An In-Context Learning Approach for Pure Exploration

    Published: 6/6/2025
20. How Bidirectionality Helps Language Models Learn Better via Dynamic Bottleneck Estimation

    Published: 6/6/2025