Best AI papers explained

550 Episodo

1. Past-Token Prediction for Long-Context Robot Policies

   Publicado: 20/5/2025
2. Recovering Coherent Event Probabilities from LLM Embeddings

   Publicado: 20/5/2025
3. Systematic Meta-Abilities Alignment in Large Reasoning Models

   Publicado: 20/5/2025
4. Predictability Shapes Adaptation: An Evolutionary Perspective on Modes of Learning in Transformers

   Publicado: 20/5/2025
5. Efficient Exploration for LLMs

   Publicado: 19/5/2025
6. Rankers, Judges, and Assistants: Towards Understanding the Interplay of LLMs in Information Retrieval Evaluation

   Publicado: 18/5/2025
7. Bayesian Concept Bottlenecks with LLM Priors

   Publicado: 17/5/2025
8. Transformers for In-Context Reinforcement Learning

   Publicado: 17/5/2025
9. Evaluating Large Language Models Across the Lifecycle

   Publicado: 17/5/2025
10. Active Ranking from Human Feedback with DopeWolfe

    Publicado: 16/5/2025
11. Optimal Designs for Preference Elicitation

    Publicado: 16/5/2025
12. Dual Active Learning for Reinforcement Learning from Human Feedback

    Publicado: 16/5/2025
13. Active Learning for Direct Preference Optimization

    Publicado: 16/5/2025
14. Active Preference Optimization for RLHF

    Publicado: 16/5/2025
15. Test-Time Alignment of Diffusion Models without reward over-optimization

    Publicado: 16/5/2025
16. Test-Time Preference Optimization: On-the-Fly Alignment via Iterative Textual Feedback

    Publicado: 16/5/2025
17. GenARM: Reward Guided Generation with Autoregressive Reward Model for Test-time Alignment

    Publicado: 16/5/2025
18. Advantage-Weighted Regression: Simple and Scalable Off-Policy RL

    Publicado: 16/5/2025
19. Can RLHF be More Efficient with Imperfect Reward Models? A Policy Coverage Perspective

    Publicado: 16/5/2025
20. Transformers can be used for in-context linear regression in the presence of endogeneity

    Publicado: 15/5/2025