Learning Adaptive Parallel Execution for Efficient Code Localization
This addresses the bottleneck of redundant tool invocations in automated software development pipelines, offering a cost-effective solution for code localization.
The paper tackled the problem of inefficient parallel execution in code localization, where current agents have a 34.9% redundant invocation rate, and proposed FuseSearch, which achieved SOTA-level performance (84.7% file-level and 56.4% function-level F1 scores) with a 93.6% speedup, using 67.7% fewer turns and 68.9% fewer tokens.
Code localization constitutes a key bottleneck in automated software development pipelines. While concurrent tool execution can enhance discovery speed, current agents demonstrate a 34.9\% redundant invocation rate, which negates parallelism benefits. We propose \textbf{FuseSearch}, reformulating parallel code localization as a \textbf{joint quality-efficiency optimization} task. Through defining \textbf{tool efficiency} -- the ratio of unique information gain to invocation count -- we utilize a two-phase SFT and RL training approach for learning adaptive parallel strategies. Different from fixed-breadth approaches, FuseSearch dynamically modulates search breadth according to task context, evolving from exploration phases to refinement stages. Evaluated on SWE-bench Verified, FuseSearch-4B achieves SOTA-level performance (84.7\% file-level and 56.4\% function-level $F_1$ scores) with 93.6\% speedup, utilizing 67.7\% fewer turns and 68.9\% fewer tokens. Results indicate that efficiency-aware training naturally improves quality through eliminating noisy redundant signals, enabling high-performance cost-effective localization agents.