FlowEvo：通过工作流与可执行技能的协同进化实现自进化智能体

摘要

大语言模型智能体可以通过在推理时构建工作流来适应复杂任务，但单个回合中发现的过程通常会在执行后被丢弃。现有技能库提供可复用的可执行例程，但通常离线构建，且不能从智能体自身的工作流中增长。我们提出 FlowEvo，一种无需训练的框架，让工作流和技能在推理时共同演化。FlowEvo 将成功的工作流整合为可调用的技能，将其存储在持久化技能库中，并通过直接执行或作为构建新工作流的上下文来使用检索到的技能。它还跟踪每个技能的下游效用，并抑制导致负迁移的技能。在共享 GPT-4o-mini 主干下，FlowEvo 在 ALFWorld、HumanEval、MBPP、GSM8K 和 MATH-500 的完整标准划分上，于 8 个基线中取得最高准确率。在 ALFWorld 上，FlowEvo 达到 85.6%，比最强基线高出 26.4 个百分点，同时仅使用约三分之一的 token 数。在涵盖 7B 到 671B 参数的 10 个基座模型上，FlowEvo 在 50 个模型-数据集对比中优于 ExpeL 49 个。代码见 https://github.com/DEFENSE-SEU/FlowEvo。

English

Large language model agents can adapt to complex tasks by constructing workflows at inference time, but procedures discovered in one episode are usually discarded after execution. Existing skill libraries provide reusable executable routines, but are typically assembled offline and do not grow from the agent's own workflows. We introduce FlowEvo, a training-free framework in which workflows and skills co-evolve at inference time. FlowEvo compiles successful workflows into callable skills, stores them in a persistent bank, and uses retrieved skills either through direct execution or as context for constructing new workflows. It also tracks each skill's downstream utility and suppresses skills that cause negative transfer. Using a shared GPT-4o-mini backbone, FlowEvo achieves the highest accuracy among 8 baselines on the full standard splits of ALFWorld, HumanEval, MBPP, GSM8K, and MATH-500. On ALFWorld, it reaches 85.6%, 26.4 points above the strongest baseline, while using roughly one third as many tokens. Across 10 base models spanning 7B to 671B parameters, FlowEvo outperforms ExpeL in 49 of 50 model-dataset comparisons. Code is available at https://github.com/DEFENSE-SEU/FlowEvo.