Steer, Don't Solve: Training Small Critic Models for Large Code Agents.

Shubham Gandhi Yiqing Xie Atharva Naik Ruichen Zhu Carolyn P. Rosé Steer, Don't Solve: Training Small Critic Models for Large Code Agents. 2026 June abs/2606.21811 CoRR https://doi.org/10.48550/arXiv.2606.21811 db/journals/corr/corr2606.html#abs-2606-21811 streams/journals/corr