Best AI papers explained

456 Episodes

1. Test-Time Alignment of Diffusion Models without reward over-optimization

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

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

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

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

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

   Published: 5/15/2025
7. Bayesian Concept Bottlenecks with LLM Priors

   Published: 5/15/2025
8. In-Context Parametric Inference: Point or Distribution Estimators?

   Published: 5/15/2025
9. Enough Coin Flips Can Make LLMs Act Bayesian

   Published: 5/15/2025
10. Bayesian Scaling Laws for In-Context Learning

    Published: 5/15/2025
11. Posterior Mean Matching Generative Modeling

    Published: 5/15/2025
12. Can Generative AI Solve Your In-Context Learning Problem? A Martingale Perspective

    Published: 5/15/2025
13. Dynamic Search for Inference-Time Alignment in Diffusion Models

    Published: 5/15/2025
14. Is In-Context Learning in Large Language Models Bayesian? A Martingale Perspective

    Published: 5/12/2025
15. Leaked Claude Sonnet 3.7 System Instruction tuning

    Published: 5/12/2025
16. Converging Predictions with Shared Information

    Published: 5/11/2025
17. Test-Time Alignment Via Hypothesis Reweighting

    Published: 5/11/2025
18. Rethinking Diverse Human Preference Learning through Principal Component Analysis

    Published: 5/11/2025
19. Active Statistical Inference

    Published: 5/10/2025
20. Data Mixture Optimization: A Multi-fidelity Multi-scale Bayesian Framework

    Published: 5/10/2025