AI研究論文每日精選

CLEAR: Character Unlearning in Textual and Visual Modalities

AutoKaggle: A Multi-Agent Framework for Autonomous Data Science Competitions

SocialGPT: Prompting LLMs for Social Relation Reasoning via Greedy Segment Optimization

Flow-DPO: Improving LLM Mathematical Reasoning through Online Multi-Agent Learning

Can Language Models Replace Programmers? REPOCOD Says 'Not Yet'

OpenWebVoyager: Building Multimodal Web Agents via Iterative Real-World Exploration, Feedback and Optimization

Precise and Dexterous Robotic Manipulation via Human-in-the-Loop Reinforcement Learning

Mind Your Step (by Step): Chain-of-Thought can Reduce Performance on Tasks where Thinking Makes Humans Worse

Task Vectors are Cross-Modal

ShadowKV: KV Cache in Shadows for High-Throughput Long-Context LLM Inference

Robots Pre-train Robots: Manipulation-Centric Robotic Representation from Large-Scale Robot Dataset

Zero-Shot Dense Retrieval with Embeddings from Relevance Feedback

Accelerating Direct Preference Optimization with Prefix Sharing

RARe: Retrieval Augmented Retrieval with In-Context Examples

Measuring memorization through probabilistic discoverable extraction