Best AI papers explained

427 Episodes

1. Almost Surely Safe LLM Inference-Time Alignment

   Published: 5/23/2025
2. Survey of In-Context Learning Interpretation and Analysis

   Published: 5/23/2025
3. From Decoding to Meta-Generation: Inference-time Algorithms for Large Language Models

   Published: 5/23/2025
4. LLM In-Context Learning as Kernel Regression

   Published: 5/23/2025
5. Where does In-context Learning Happen in Large Language Models?

   Published: 5/23/2025
6. Auto-Differentiating Any LLM Workflow: A Farewell to Manual Prompting

   Published: 5/22/2025
7. metaTextGrad: Learning to learn with language models as optimizers

   Published: 5/22/2025
8. Semantic Operators: A Declarative Model for Rich, AI-based Data Processing

   Published: 5/22/2025
9. Isolated Causal Effects of Language

   Published: 5/22/2025
10. Sleep-time Compute: Beyond Inference Scaling at Test-time

    Published: 5/22/2025
11. J1: Incentivizing Thinking in LLM-as-a-Judge

    Published: 5/22/2025
12. ShiQ: Bringing back Bellman to LLMs

    Published: 5/22/2025
13. Policy Learning with a Natural Language Action Space: A Causal Approach

    Published: 5/22/2025
14. Multi-Objective Preference Optimization: Improving Human Alignment of Generative Models

    Published: 5/22/2025
15. End-to-End Learning for Stochastic Optimization: A Bayesian Perspective

    Published: 5/21/2025
16. TEXTGRAD: Automatic Differentiation via Text

    Published: 5/21/2025
17. Steering off Course: Reliability Challenges in Steering Language Models

    Published: 5/20/2025
18. Past-Token Prediction for Long-Context Robot Policies

    Published: 5/20/2025
19. Recovering Coherent Event Probabilities from LLM Embeddings

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

    Published: 5/20/2025