Best AI papers explained

437 Episodo

1. Emergent Misalignment: Narrow finetuning can produce broadly misaligned LLMs

   Publicado: 11/6/2025
2. Agentic Supernet for Multi-agent Architecture Search

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

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

   Publicado: 10/6/2025
5. LLMs Get Lost In Multi-Turn Conversation

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

   Publicado: 8/6/2025
7. General Agents Need World Models

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

   Publicado: 7/6/2025
9. Decisions With Algorithms

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

    Publicado: 6/6/2025
11. Conformal Arbitrage for LLM Objective Balancing

    Publicado: 6/6/2025
12. Simulation-Based Inference for Adaptive Experiments

    Publicado: 6/6/2025
13. Agents as Tool-Use Decision-Makers

    Publicado: 6/6/2025
14. Quantitative Judges for Large Language Models

    Publicado: 6/6/2025
15. Self-Challenging Language Model Agents

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

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

    Publicado: 6/6/2025
18. A Closer Look at Bias and Chain-of-Thought Faithfulness of Large (Vision) Language Models

    Publicado: 5/6/2025
19. Simplifying Bayesian Optimization Via In-Context Direct Optimum Sampling

    Publicado: 5/6/2025
20. Bayesian Teaching Enables Probabilistic Reasoning in Large Language Models

    Publicado: 5/6/2025