探索时持异议，提交时持共识：面向软件智能体的路由引导测试时缩放

中文摘要 AI 辅助

软件工程智能体通过漫长且随机的工具使用轨迹解决仓库级任务，多次尝试常能找到单次运行遗漏的修复方案。由于补丁缺乏规范的答案形式，且来自同一前缀的兄弟动作存在相关性，测试时缩放颇具难度。我们研究能否利用原生混合专家（MoE）路由轨迹进行引导和选择，无需外部判断或选择时的测试执行。分析表明，路由提供了稳健的行为角色信号；令牌粒度读数与决策匹配的对比集可将其转化为有效控制。因此，我们提出Risa（路由引导的转向与仲裁）：在轨迹内部，路由在补丁提交过程中鼓励多样化探索与受控收敛；在单独采样的轨迹之间，信息补丁位置的共识可选择最终候选方案。我们在SWE-bench Verified上，使用开放权重的稀疏MoE智能体，跨规模和推理努力设置进行评估。Risa的路由仲裁将宏观平均解决率从均匀采样下的44.9%提升至gpt-oss系列的48.2%，无需答案字符串匹配即可达到文本共识的效果，且该方法可迁移至Qwen3.6，在完整500任务基准上，其性能优于均匀选择并达到文本共识水平。

英文摘要

Software-engineering agents solve repository-level tasks through long, stochastic tool-use trajectories, and repeated attempts often find fixes missed by one run. Test-time scaling is difficult because patches lack canonical answer forms, while sibling actions from a shared prefix are correlated. We study whether native MoE router traces can guide steering and selection without an external judge or selection-time test execution. Our analysis shows that routing provides a robust behavioral role signal; token-granular readouts and decision-matched comparison sets turn it into effective control. We therefore introduce Risa (Routing-Informed Steering and Arbitration): within trajectories, routing encourages diverse exploration and controlled convergence during patch commitment; across separately sampled trajectories, agreement at informative patch positions selects a final candidate. We evaluate on SWE-bench Verified using open-weight sparse MoE agents across scales and reasoning-effort settings. Risa's routing arbitration raises the macro-average resolved rate from 44.9% under uniform sampling to 48.2% on the gpt-oss family, matching text consensus without answer-string matching, and it transfers to Qwen3.6, where it improves on uniform choice and matches text consensus on the full 500-task benchmark.

发表机构

• Fudan University（复旦大学）
• Shanghai Innovation Institute（上海创新研究院）

机构由 AI 辅助整理，请以论文原文为准。