Best AI papers explained

456 Episodes

1. Trace is the Next AutoDiff: Generative Optimization with Rich Feedback, Execution Traces, and LLMs

   Published: 5/24/2025
2. Adaptive Inference-Time Compute: LLMs Can Predict if They Can Do Better, Even Mid-Generation

   Published: 5/24/2025
3. Prompts from Reinforcement Learning (PRL)

   Published: 5/24/2025
4. Logits are All We Need to Adapt Closed Models

   Published: 5/24/2025
5. Large Language Models Are (Bayesian) Latent Variable Models: Explaining and Finding Good Demonstrations for In-Context Learning

   Published: 5/23/2025
6. Inference-Time Intervention: Eliciting Truthful Answers from a Language Model

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

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

   Published: 5/23/2025
9. Personalizing LLMs via Decode-Time Human Preference Optimization

   Published: 5/23/2025
10. Almost Surely Safe LLM Inference-Time Alignment

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

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

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

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

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

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

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

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

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

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

    Published: 5/22/2025