arXiv 2610.27450 · cs.AI · 2026-10-20
Ask to Distill: Learning to Synthesize Data for On-Policy Distillation
Jiashuo Sun, Hao Zhang, Xueqiang Xu, Heng Wang, Yixuan Xie, Pengcheng Jiang, Zhenrui Yue, Jiajun Fan and 4 more
University of Illinois at Urbana-Champaign
In latest · #342 Added on 2026-10-25
One-line takeaway
UIUC's Ask-to-Distill learns to synthesize questions for on-policy distillation, scoring question value with a reference-free distributional surrogate from a frozen student-teacher pair and optimizing a student-initialized synthesizer on it.
UIUC 的 Ask-to-Distill 学习为在线策略蒸馏合成训练题：把题目当作暴露教师知识的载体，用冻结的师生模型算出无需标准答案的题目价值替代指标，据此优化出题器。
Read on arXiv PDF Hugging Face discussion Code
Topics & disciplines
More on Efficient training & inference
- The Teacher Is a Direction, Not a Destination: Extrapolating RL-Induced Representation Residuals in On-Policy Distillation ▲511
- Scaling Properties of Same-Family On-Policy Distillation ▲327
- Beyond Domain-Routed Supervision: Cross-Teacher Guarding for Multi-Teacher On-Policy Distillation ▲250
- On-Policy or Off-Policy Learning? A Systematic Study of Distillation Dynamics ▲205
- Beyond Teacher Assignment: Domain-Normalized Multi-Teacher On-Policy Distillation ▲181
More on LLMs
- Jeva: System One Models from First Principles 2026-10-13 · ▲371
- Post-Training Leaves Behavioral Shadows on Unrelated Decisions 2026-09-24 · ▲273
- WeVisOne-VL: Unified Visual Perception from Objects to Pixels 2026-10-22 · ▲247
- Thinking Inertia: LLMs Keep Thinking When Told Not To 2026-10-08 · ▲241
- OmniEdu: Open Foundation Models for Learning and Teaching 2026-09-19 · ▲240
More from University of Illinois at Urbana-Champaign
About this data
Metadata from the arXiv API (verified); upvotes from Hugging Face Daily Papers; citations from Semantic Scholar. Last refreshed 2026-10-26.
The takeaway is written by AIMap from the abstract; the abstract itself is not reproduced — read the original.