Best AI papers explained

515 Épisodes

1. LoRA Without Regret

   Publié: 01/10/2025
2. Actor-Critic without Actor: Critic-Guided Denoising for RL

   Publié: 29/09/2025
3. DELTA-Code: How Does RL Unlock and Transfer New Programming Algorithms in LLMs?

   Publié: 29/09/2025
4. Linear Transformers Implicitly Discover Unified Numerical Algorithms

   Publié: 29/09/2025
5. Regularizing Extrapolation in Causal Inference

   Publié: 27/09/2025
6. DoubleGen - Debiased Generative Modeling of Counterfactuals

   Publié: 27/09/2025
7. What Characterizes Effective Reasoning? Revisiting Length, Review, and Structure of CoT

   Publié: 27/09/2025
8. Compute as Teacher: Turning Inference Compute Into Reference-Free Supervision

   Publié: 27/09/2025
9. Learning without training: The implicit dynamics of in-context learning

   Publié: 24/09/2025
10. Does Reinforcement Learning Really Incentivize Reasoning Capacity in LLMs Beyond the Base Model

    Publié: 24/09/2025
11. Open Problems in Mechanistic Interpretability

    Publié: 21/09/2025
12. Maestro: Joint Graph & Config Optimization for Reliable AI Agents

    Publié: 21/09/2025
13. Thought Anchors: Which LLM Reasoning Steps Matter?

    Publié: 21/09/2025
14. Sample Complexity and Representation Ability of Test-time Scaling Paradigms

    Publié: 09/09/2025
15. RL's Razor: Why Online RL Forgets Less

    Publié: 07/09/2025
16. Why Language Models Hallucinate

    Publié: 06/09/2025
17. ALFA: Aligning LLMs to Ask Good Questions A Case Study in Clinical Reasoning

    Publié: 06/09/2025
18. Sample Efficient Preference Alignment in LLMs via Active Exploration

    Publié: 06/09/2025
19. Adventures in Demand Analysis Using AI

    Publié: 04/09/2025
20. Memento: Fine-tuning LLM Agents without Fine-tuning LLMs

    Publié: 01/09/2025