Best AI papers explained

522 Afleveringen

1. Personalized reasoning: just-in-time personalization and why LLMs fail at it

   Gepubliceerd: 5-10-2025
2. Prompt Curriculum Learning for Efficient LLM Post-Training

   Gepubliceerd: 5-10-2025
3. Personalizing Reinforcement Learning from Human Feedback with Variational Preference Learning

   Gepubliceerd: 4-10-2025
4. Enhancing Personalized Multi-Turn Dialogue with Curiosity Reward

   Gepubliceerd: 4-10-2025
5. Learning to summarize user information for personalized reinforcement learning from human feedback

   Gepubliceerd: 4-10-2025
6. Distributional Preference Learning: Understanding and Accounting for Hidden Context in RLHF

   Gepubliceerd: 3-10-2025
7. LIMI: Less is More for Agency

   Gepubliceerd: 1-10-2025
8. LoRA Without Regret

   Gepubliceerd: 1-10-2025
9. Actor-Critic without Actor: Critic-Guided Denoising for RL

   Gepubliceerd: 29-9-2025
10. DELTA-Code: How Does RL Unlock and Transfer New Programming Algorithms in LLMs?

    Gepubliceerd: 29-9-2025
11. Linear Transformers Implicitly Discover Unified Numerical Algorithms

    Gepubliceerd: 29-9-2025
12. Regularizing Extrapolation in Causal Inference

    Gepubliceerd: 27-9-2025
13. DoubleGen - Debiased Generative Modeling of Counterfactuals

    Gepubliceerd: 27-9-2025
14. What Characterizes Effective Reasoning? Revisiting Length, Review, and Structure of CoT

    Gepubliceerd: 27-9-2025
15. Compute as Teacher: Turning Inference Compute Into Reference-Free Supervision

    Gepubliceerd: 27-9-2025
16. Learning without training: The implicit dynamics of in-context learning

    Gepubliceerd: 24-9-2025
17. Does Reinforcement Learning Really Incentivize Reasoning Capacity in LLMs Beyond the Base Model

    Gepubliceerd: 24-9-2025
18. Open Problems in Mechanistic Interpretability

    Gepubliceerd: 21-9-2025
19. Maestro: Joint Graph & Config Optimization for Reliable AI Agents

    Gepubliceerd: 21-9-2025
20. Thought Anchors: Which LLM Reasoning Steps Matter?

    Gepubliceerd: 21-9-2025