Speculative Macro Commit for Faster Tool-Using Agents

论文概要

研究领域: AI/ML 作者: Zeyu Liu, Souvik Kundu, Peter A. Beerel 发布时间: 2026-09-06 arXiv: 2509.00003

中文摘要

工具使用型LLM智能体不仅在模型推理上消耗挂钟时间，还在串行的动作-观察轮次中耗费时间，其中每个工具调用、环境转换和观察都可能延迟后续决策。我们提出了Speculative Macro Commit (SMC)，一种用于双层智能体系统的运行时机制：一个大型权威执行模型生成官方轨迹，而一个更快的推测起草模型在隔离的环境快照上持续预测并执行未来动作链。SMC从训练轨迹中挖掘重复的多动作骨架，并将其存储在宏库中，用于与运行时起草模型预测的动作链进行匹配。当执行模型的下一个工具调用与第一个起草动作匹配时，SMC将剩余预执行的草稿步骤及其观察结果提交到官方轨迹。使用Qwen3.5-27B INT4作为权威执行模型、Qwen3.5-4B作为推测起草模型，SMC在保持顺序智能体整体准确率的同时，在τ²-Bench Telecom子集上将延迟比Speculative Actions (SA)基线降低了10.23%，比顺序执行降低了18.59%。在AppWorld上，SMC比SA基线减少了7.7%的挂钟时间，比顺序执行减少了44.9%，任务完成率仅有小幅下降。总体而言，SMC提供了一种实用的方法来重用多步推测执行，并将智能体延迟降低到超越单步推测动作的水平。

原文摘要

Tool-using LLM agents spend wall-clock time not only on model inference but also in serial action--observation turns, where each tool call, environment transition, and observation can delay subsequent decisions. We introduce Speculative Macro Commit (SMC), a runtime mechanism for a two-tier agent system: a large authoritative actor model produces the official trajectory, while a faster speculative drafter model continuously predicts and executes future action chains on an isolated environment snapshot. SMC mines recurring multi-action skeletons from training traces and stores them in a macro library used to match against action chains predicted by the drafter at runtime. When the actor's next tool call matches the first drafted action, SMC commits the remaining pre-executed draft steps, to...

#论文 #arXiv #AI #小凯