Best AI papers explained

442 Episodes

1. Leaked Claude Sonnet 3.7 System Instruction tuning

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

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

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

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

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

   Published: 5/10/2025
7. AI-Powered Bayesian Inference

   Published: 5/10/2025
8. Can Unconfident LLM Annotations Be Used for Confident Conclusions?

   Published: 5/9/2025
9. Predictions as Surrogates: Revisiting Surrogate Outcomes in the Age of AI

   Published: 5/9/2025
10. Learn then Test: Calibrating Predictive Algorithms to Achieve Risk Control

    Published: 5/9/2025
11. How to Evaluate Reward Models for RLHF

    Published: 5/9/2025
12. LLMs as Judges: Survey of Evaluation Methods

    Published: 5/9/2025
13. The Alternative Annotator Test for LLM-as-a-Judge: How to Statistically Justify Replacing Human Annotators with LLMs

    Published: 5/9/2025
14. Limits to scalable evaluation at the frontier: LLM as Judge won’t beat twice the data

    Published: 5/9/2025
15. Stratified Prediction-Powered Inference for Hybrid Language Model Evaluation

    Published: 5/9/2025
16. Accelerating Unbiased LLM Evaluation via Synthetic Feedback

    Published: 5/9/2025
17. Prediction-Powered Statistical Inference Framework

    Published: 5/9/2025
18. Optimizing Chain-of-Thought Reasoners via Gradient Variance Minimization in Rejection Sampling and RL

    Published: 5/9/2025
19. RM-R1: Reward Modeling as Reasoning

    Published: 5/9/2025
20. Reexamining the Aleatoric and Epistemic Uncertainty Dichotomy

    Published: 5/8/2025