arXiv cs.AI· Jizhe Zhang, Ang Li, Chenheng Zhang, Yisen Wang·· 11 小时前AI 评分47
SumOne：为即插即用智能体记忆打造的高效摘要器
SUMONE: A Simple and Efficient Summarizer for Plug-and-Play Agent Memory
AI 导读
SumOne 是一个独立训练的摘要器，通过掩码重建让冻结的重建器还原原始轨迹中被随机遮蔽的部分，结合长度惩罚构成任务无关奖励，用 GRPO 优化，无需参考摘要或下游任务执行。
正文 · 原文
Abstract:Long-horizon language-model agents accumulate increasingly large interaction histories, making context compaction essential for efficient execution. Existing learned compaction methods often rely on downstream task outcomes or solver behavior, coupling training to specific tasks or agents. We introduce SumOne, a standalone summarizer trained to preserve information in agent histories through masked reconstruction. Given a summary of the expired history, a frozen reconstructor recovers randomly masked portions of the original trajectory; reconstruction quality, combined with a length penalty, provides a task-agnostic reward for optimizing the summarizer with GRPO. This yields self-supervision directly from raw trajectories, without reference summaries or downstream task execution. At inference time, only the lightweight summarizer is retained, enabling plug-and-play context compression for otherwise unchanged agents. We conduct a systematic evaluation of SumOne in long-horizon search and software-engineering tasks. SumOne reduces estimated inference cost by approximately 78\% on DeepSearchQA and 75\% on SWE-bench Verified relative to operating with full context, with absolute decreases of 0.04. The same checkpoint also transfers across downstream models and agent harnesses without additional fine-tuning.
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2611.06445 [cs.AI] |
| (or arXiv:2611.06445v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2611.06445 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Jizhe Zhang [view email]
[v1]
Wed, 4 Nov 2026 12:49:07 UTC (1,488 KB)
来源：arXiv cs.AI · arxiv.org