Critic models and SFT corpora from 'Steer, Don't Solve: Training Small Critic Models for Large Code Agents' (arXiv 2606.21811).
-
Steer, Don't Solve: Training Small Critic Models for Large Code Agents
Paper • 2606.21811 • Published • 1 -
code-critic-model/Qwen3-8B-Critic-SFT
Text Generation • 1B • Updated • 338 -
code-critic-model/Qwen3-4B-Critic-SFT
Text Generation • 4B • Updated • 378 -
code-critic-model/Qwen3-4B-Critic-SFT-DPO
Text Generation • 4B • Updated • 351