#on-policy distillation
On-Policy Distillation: 1 AI articles covering on-policy distillation news, analysis, and research
Articles
Improving OCR Faithfulness via Gated and Attenuated On-PolicyNEW⭐8
GAD-RL adaptively gates and attenuates teacher supervision during on-policy distillation, improving OCR faithfulness and achieving strong CHAOS-Bench and OmniDo...
