arXiv：cs.AI· Jan-Peter Franke·· 3 小时前
面向工具调用智能体的智能体控制遗忘：可逆上下文精简实践
Agent-Controlled Forgetting for Tool-Using Agents: Reversible Context Curation in Practice
AI 导读
研究提出"智能体控制遗忘"方法：执行模型自己选定已观察到的工具结果，用短注释替换原位置内容，并将原始结果存入可恢复归档，且不触动用户指令与助手消息。
正文
Abstract:Tool-using agents repeatedly carry observations whose useful content can be much smaller than their original payload. We study agent-controlled forgetting: the acting model selects previously observed tool results, replaces each with a short note at its original position, and retains the exact original in a recoverable archive. A Python harness exposes batch archival and explicit recovery without task-specific model training, while protecting user instructions and assistant messages from these operations. In an exploratory OpenTelemetry debugging case followed by an unrelated implementation task, the method ended with 231,951 provider-reported prompt tokens versus 912,492 under retained history, used 50% fewer cumulative input tokens, and had an estimated API cost of USD 1.28-1.44 versus approximately USD 4.38. Both arms passed the two-case primary behavioral oracle; neither fully satisfied the follow-up evaluation. The method made more requests and took 17% longer. A contrasting application-development pair produced no context or cost saving, and an earlier continuation exhibited lower manually assessed quality despite reduced context. These observations demonstrate substantial resource savings in noisy tool-use trajectories and identify workload dependence as a central consideration for reversible context management.
| Comments: | 13 pages, 1 figure. Code and research artifacts: this https URL |
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.10590 [cs.AI] |
| (or arXiv:2610.10590v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.10590 arXiv-issued DOI via DataCite |
Submission history
From: Jan-Peter Franke [view email]
[v1]
Tue, 6 Oct 2026 18:13:48 UTC (38 KB)
来源：arXiv：cs.AI · arxiv.org