arXiv cs.AI· Peiwen Li (Yale University), Harshit Verma (Yale University), Sophia Zhou (Yale University), Haiwen Wang (Yale University), Ngoc Bui (Yale University), Heng Chang (Tsinghua University), Fei Fei (Goodvision AI), David van Dijk (Yale University), Rex Ying (Yale University)·· 1 天前AI 评分34
CoReMem：面向连续编码任务的内化成本记忆框架
CoReMem: Cost-Internalizing Memory for Successive Coding Tasks with Prospective Valuation
AI 导读
CoReMem 是一个内化成本记忆框架，通过前瞻性片段估值比较下一任务有无继承轨迹片段的表现，引导构建以仓库位置和可执行检查为锚点的紧凑记忆。在 ChainSWE 上，27B actor 的解决率从 21.0% 提升至 25.7%，步数减少 17%，含记忆模块调用的总 cache-aware token 减少 11%。
正文 · 原文
Abstract:Coding agents tackle successive tasks on an evolving repository, carrying forward edits and writing memory before the next task is known. Selecting memory for unknown next tasks requires balancing future usefulness against the costs of writing and reading it. The actor must also decide when acting on available memory saves effort as the task unfolds. We introduce Coding-Repository Memory (CoReMem), a cost-internalizing framework that guides memory design and actor adaptation through measured memory value on the next task. Prospective segment valuation compares runs of the next task with and without individual trajectory segments from the same inherited repository state, guiding a compact memory built around anchors: repository locations and executable checks. Actor adaptation estimates on-policy memory value within rollout groups with and without memory, using a reward that prioritizes resolution and penalizes step count in successful runs. Provenance-gated value credit assigns the additional value term to steps identified as using memory-supplied anchors, while every step retains outcome credit. On ChainSWE, CoReMem raises the 27B actor's resolve rate from 21.0% to 25.7%, with 17% fewer steps and 11% fewer total cache-aware tokens, including memory-module calls, relative to the untrained actor model without memory.
| Subjects: | Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Software Engineering (cs.SE) |
| Cite as: | arXiv:2611.17316 [cs.AI] |
| (or arXiv:2611.17316v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2611.17316 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Peiwen Li [view email]
[v1]
Thu, 12 Nov 2026 18:58:22 UTC (842 KB)
来源：arXiv cs.AI · arxiv.org