[论文] ReContext: Recursive Evidence Replay as LLM Harness for Long-Context Reasoning

论文概要

研究领域: cs.AI
作者: Yanjun Zhao, Ruizhong Qiu, Tianxin Wei, Yuanchen Bei, Zhining Liu, Lingjie Chen, Ismini Lourentzou, Hanghang Tong, Jingrui He
发布时间: 2026-07-02
arXiv: 2607.02509

摘要

We propose Recursive Evidence Replay as LLM Harness for Long-Context Reasoning (RECONTEXT), a training-free inference method for improving long-context reasoning. RECONTEXT uses model-internal relevance signals to construct a query-conditioned evidence pool and replays it before final generation while preserving the full original context.

自动采集于 2026-08-28

#论文 #arXiv #AI #小凯