arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2610.20116cs.AI
收藏

最大化上下文压缩的透明度可提升轻量编码智能体对事实和承诺的保留能力

Maximising Transparency in Context Compaction Improves Retention of Facts and Commitments for Lightweight Coding Agents

Center for Human Nature, Artificial Intelligence, and Neuroscience (CHAIN)(人类自然、人工智能与神经科学中心);Hokkaido University(北海道大学)

Felix Woolford

AI总结:

本文提出transparent-compact方法,对比Pi默认压缩与pi-vcc,在含Qwen3.8-27B的四模型基准测试中,该方法提升轻量编码智能体的事实与承诺保留能力,减少自信错误。

摘要:

长时编码智能体会话会超出模型的上下文窗口,因此智能体会进行压缩:用更短的内容替换旧的历史记录,通常是模型生成的摘要。摘要会进行选择和改写,而智能体无法看到丢失的内容。现有研究已探索了多种压缩技术,但对其在轻量模型(包括本地部署或用于高工作量的模型)上的有效性关注相对较少。本文提出了transparent-compact,这是一种针对开源Pi编码智能体的简单确定性方法,专为这类模型设计,目的是最大化智能体对自身压缩上下文的视图透明度。它保证上下文中的用户和助手文本不会被模型或算法修改,并明确标记压缩已剥离信息的位置。我们将其与Pi默认的模型生成摘要压缩、以及pi-vcc(一种类似方法但不逐字保留文本的流行确定性方法)进行了比较。我们在四个不同性能水平的模型上运行基准测试,重点是Qwen3.8-27B——一款消费级硬件可本地运行的顶级轻量模型。在探测承诺(仅存在于对话中的用户设定规则)和事实回忆的数据集上,transparent-compact的表现与两种替代方法相当或更优。尤其在避免自信错误(即断言虚假信息而非识别未回忆内容)方面,它展现出优势;即使在前沿模型上,其优势依然存在但幅度较小。我们认为,transparent-compact可让轻量模型无需推理或规避压缩丢失的内容,从而提升信息的回忆和应用效果,且无需对会话历史进行任务或上下文特定的结构化处理。

英文摘要:

Long coding agent sessions outgrow the model's context window, so agents compact: older history is replaced by something shorter, usually a model-written summary. A summary selects and rewords, and the agent cannot see what was lost. Existing work has explored a range of compaction techniques, but comparatively little attention has been paid to their effectiveness with lightweight models, including those deployed locally or used for high-volume workloads. Here, we present transparent-compact, a simple deterministic method for the open-source Pi coding agent, designed specifically for use with such models and with the purpose of maximising the transparency of the agent's view of its own compacted context. It guarantees that user and assistant prose in its context is unchanged by a model or algorithm, and explicitly flags where compaction has stripped information. We compare it with Pi's default model-written summary compaction and with pi-vcc, a popular deterministic method with similar methods but without verbatim retention of prose. Our benchmarks run on four models of varying performance levels, with emphasis on Qwen3.8-27B: a lightweight model at the top end of what consumer hardware can serve locally. On sets probing commitments (rules the user set which exist only in the conversation) and recall of facts, transparent-compact matched or outperformed both alternatives. In particular, it showed an advantage in avoiding confident blunders, i.e. assertion of false information as opposed to recognizing that something was not recalled. Even with a frontier model, its advantage was present but smaller. We argue that transparent-compact spares lightweight models from reasoning or searching their way around what compaction lost, and results in improved recall and application of information as a result, without the need for task or context-specific structuring of the session history.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑