Best AI papers explained

515 Episodo

1. AI Agents Need Authenticated Delegation

   Publicado: 25/6/2025
2. Probabilistic Modelling is Sufficient for Causal Inference

   Publicado: 25/6/2025
3. Not All Explanations for Deep Learning Phenomena Are Equally Valuable

   Publicado: 25/6/2025
4. e3: Learning to Explore Enables Extrapolation of Test-Time Compute for LLMs

   Publicado: 17/6/2025
5. Extrapolation by Association: Length Generalization Transfer in Transformers

   Publicado: 17/6/2025
6. Uncovering Causal Hierarchies in Language Model Capabilities

   Publicado: 17/6/2025
7. Generalization or Hallucination? Understanding Out-of-Context Reasoning in Transformers

   Publicado: 17/6/2025
8. Improving Treatment Effect Estimation with LLM-Based Data Augmentation

   Publicado: 17/6/2025
9. LLM Numerical Prediction Without Auto-Regression

   Publicado: 17/6/2025
10. Self-Adapting Language Models

    Publicado: 17/6/2025
11. Why in-context learning models are good few-shot learners?

    Publicado: 17/6/2025
12. Take Caution in Using LLMs as Human Surrogates: Scylla Ex Machina∗

    Publicado: 14/6/2025
13. The Logic of Machines: The AI Reasoning Debate

    Publicado: 12/6/2025
14. Layer by Layer: Uncovering Hidden Representations in Language Models

    Publicado: 12/6/2025
15. Causal Attribution Analysis for Continuous Outcomes

    Publicado: 12/6/2025
16. Training a Generally Curious Agent

    Publicado: 12/6/2025
17. Estimation of Treatment Effects Under Nonstationarity via Truncated Difference-in-Q’s

    Publicado: 12/6/2025
18. Strategy Coopetition Explains the Emergence and Transience of In-Context Learning

    Publicado: 12/6/2025
19. Emergent Misalignment: Narrow finetuning can produce broadly misaligned LLMs

    Publicado: 11/6/2025
20. Agentic Supernet for Multi-agent Architecture Search

    Publicado: 11/6/2025