arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2602.19633cs.AI
收藏

TAPE: 语言模型代理中的工具引导自适应规划与约束执行

TAPE: Tool-Guided Adaptive Planning and Constrained Execution in Language Model Agents

Jongwon Jeong, Jungtaek Kim, Kangwook Lee

AI总结:

TAPE通过工具引导的自适应规划与约束执行方法,提升语言模型代理在复杂任务中的成功率,特别是在困难设置中表现显著优于现有框架。

摘要:

语言模型(LM)代理在解决需要与环境多次交互的任务中表现出显著的能力。然而,在单个错误往往导致不可逆失败的环境中,特别是在严格可行性约束下,它们仍然存在脆弱性。我们系统地分析了现有的代理框架,发现不完美的规划和随机执行是主要原因。为了解决这些挑战,我们提出了工具引导的自适应规划与约束执行(TAPE)。TAPE通过将多个计划聚合到图中,并利用外部求解器来识别可行路径来增强规划能力。在执行过程中,TAPE采用约束解码以减少采样噪声,同时在环境反馈偏离预期状态时自适应重新规划。在Sokoban、ALFWorld、MuSiQue和GSM8K-Hard上的实验表明,TAPE在多个任务上均优于现有框架,特别是在困难设置中表现尤为突出,平均在困难设置上将成功率提高了21.0个百分点,在较弱基础模型上平均提高了20.0个百分点。代码和数据可在此处获取。

英文摘要:

Language Model (LM) agents have demonstrated remarkable capabilities in solving tasks that require multiple interactions with the environment. However, they remain vulnerable in environments where a single error often leads to irrecoverable failure, particularly under strict feasibility constraints. We systematically analyze existing agent frameworks, identifying imperfect planning and stochastic execution as the primary causes. To address these challenges, we propose Tool-guided Adaptive Planning with constrained Execution (TAPE). TAPE enhances planning capability by aggregating multiple plans into a graph and employing an external solver to identify a feasible path. During execution, TAPE employs constrained decoding to reduce sampling noise, while adaptively re-planning whenever environmental feedback deviates from the intended state. Experiments across Sokoban, ALFWorld, MuSiQue, and GSM8K-Hard demonstrate that TAPE consistently outperforms existing frameworks, with particularly large gains on hard settings, improving success rates by 21.0 percentage points on hard settings on average, and by 20.0 percentage points for weaker base models on average. Code and data available at here.

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。

↑