Terminal-Universe：将智能体轨迹转化为可扩展终端环境

AI 摘要

Terminal-Universe 框架通过从智能体轨迹中重建可执行的工作空间，合成多样化训练任务，并利用监督微调提升智能体性能。该方法将每条轨迹转化为可复用环境，并沿广度和深度两个维度扩展任务，最终生成 37.3k 个任务环境，使 Qwen3.5-27B 在 Terminal-Bench 2.1 上单轮性能提升 11.9 分，在 EvoCode-Bench v2 上多轮性能提升 13.8 分。

正文节选

Terminal-Universe: Turning Agent Trajectories into Scalable Terminal Environments Abstract Terminal-Universe reconstructs executable workspaces from agent trajectories to synthesize diverse training tasks and improves post-training performance through supervised fine-tuning. As terminal-based code agents become prevalent, agent trajectories have accumulated at scale, while realistic, executable environments remain scarce. However, environments are what agent post-training actually requires: each

发布时间：—

抓取时间：2026-09-04 19:34

来源机构：Hugging Face