NAVER Proposes Delta Distillation Method

naver-ai · hf · 2026-07-20

NAVER AI Lab proposed **On-Policy Delta Distillation (OPD²)**, shifting the distillation objective from "imitating the teacher's output distribution" to supervising the **delta signal** between the teacher model and its base model—essentially the changes induced by instruction tuning. The authors argue that this signal transfers reasoning capabilities more directly. Experiments covering math, science, and code reasoning benchmarks demonstrate that OPD² is more stable and effective than traditional on-policy distillation, yielding stronger reasoning model performance within a shorter post-training phase. Code will be open-sourced.

Original post →

More from Research

Research channel →