SWE-RM: Execution-free Feedback For Software Engineering Agents

TL;DR

This paper introduces SWE-RM, a large-scale, execution-free reward model designed to improve software engineering agents by providing fine-grained feedback for both test-time scaling and reinforcement learning, surpassing previous methods.

Contribution

The paper develops SWE-RM, a 30-billion-parameter mixture-of-experts reward model that enhances the training and performance of SWE agents across multiple evaluation metrics.

Findings

SWE-RM significantly improves TTS and RL performance on SWE benchmarks.

It increases accuracy of existing models like Qwen3-Coder-Flash from 51.6% to 62.0%.

Achieves state-of-the-art results among open-source models.

Abstract

Execution-based feedback like unit testing is widely used in the development of coding agents through test-time scaling (TTS) and reinforcement learning (RL). This paradigm requires scalable and reliable collection of unit test cases to provide accurate feedback, and the resulting feedback is often sparse and cannot effectively distinguish between trajectories that are both successful or both unsuccessful. In contrast, execution-free feedback from reward models can provide more fine-grained signals without depending on unit test cases. Despite this potential, execution-free feedback for realistic software engineering (SWE) agents remains underexplored. Aiming to develop versatile reward models that are effective across TTS and RL, however, we observe that two verifiers with nearly identical TTS performance can nevertheless yield very different results in RL. Intuitively, TTS primarily reflects the model's ability to select the best trajectory, but this ability does not necessarily generalize to RL. To address this limitation, we identify two additional aspects that are crucial for RL training: classification accuracy and calibration. We then conduct comprehensive controlled experiments to investigate how to train a robust reward model that performs well across these metrics. In particular, we analyze the impact of various factors such as training data scale, policy mixtures, and data source composition. Guided by these investigations, we introduce SWE-RM, an accurate and robust reward model adopting a mixture-of-experts architecture with 30B total parameters and 3B activated during inference. SWE-RM substantially improves SWE agents on both TTS and RL performance. For example, it increases the accuracy of Qwen3-Coder-Flash from 51.6% to 62.0%, and Qwen3-Coder-Max from 67.0% to 74.6% on SWE-Bench Verified using TTS, achieving new state-of-the-art performance among open-source models.