Papers study multi-teacher on-policy distillation

First seen Nov 9 Updated Nov 10, 16:12 UTC Heat 13 medium evidence

Three Hugging Face papers cover on-policy distillation: P^2-MOPD uses parameter priors, Reframe targets agentic visual reasoning, and one measures tolerance to training-inference mismatch.

Timeline

Coverage

Nov 9, 00:00 UTCNov 9
P^2-MOPD: Accelerating Multi-Teacher On-Policy Distillation via Parameter Priorshuggingface.co · HF Papers · 8 comments · 3 stars
Nov 10, 00:00 UTC16h ago
Reframe: Observation-Aligned Multi-Teacher On-Policy Distillation for Agentic Visual Reasoninghuggingface.co · HF Papers · 1 comments · 14 stars
Nov 10, 00:00 UTC16h ago
How Much Training-Inference Mismatch Can On-Policy Distillation Tolerate?huggingface.co · HF Papers · 5 comments · 3 stars

Evidence

Medium evidence: several community reports point the same way, but there is no official confirmation yet.