跳到正文
HuggingFace Daily Papers(社區熱門論文)·· 2026-10-14AI 評分45

ActionOPD:蒸餾決策,而非模型 token

ActionOPD: Distilling Decisions, Not Tokens

AI 導讀

ActionOPD提出以語義動作為監督單位的動作級在策略蒸餾,讓學生模型在自身遇到的狀態下對齊教師決策。以兩名教師訓練 Qwen3-4B 和 Qwen3-8B,在四個分佈外基準取得 3.3–6.9 個百分點的平均提升。方法無需重現教師撰寫的推理或取得教師 token 機率,且在 τ^2-bench 上表現媲美離線蒸餾和 token 級 OPD。

正文

Published on Oct 14

Authors:

,

,

,

,

,

,

Abstract

Distilling capable language-model agents into smaller models requires aligning the student's behavior with the teacher's. This raises a fundamental question: at which states and at what granularity should this alignment be performed? Offline imitation provides dense but token-specific supervision and can suffer from state-distribution mismatch. Token-level on-policy distillation (OPD) addresses this mismatch but typically requires teacher token probabilities and compatible tokenization. We argue that semantic actions provide a natural unit of supervision for agent distillation. We introduce ActionOPD, an action-level on-policy distillation method motivated by maximizing expected teacher-action likelihood at student-encountered states. At each such state, ActionOPD samples actions from both the student and a black-box teacher, rewarding student actions according to their semantic agreement with the sampled teacher actions. Through policy optimization, ActionOPD aligns the student's action-level behavior with the teacher's decisions without requiring reproduction of teacher-written reasoning or access to teacher token probabilities. Training Qwen3-4B and Qwen3-8B with two teachers on diverse, curated tool-use tasks yields average gains of 3.3-6.9 percentage points across four out-of-distribution benchmarks. ActionOPD scales favorably with student size and teacher capability, without requiring the teacher and student to share a model family or tokenizer. On τ^2-bench, ActionOPD achieves competitive performance against offline distillation and token-level OPD. We further explore Self-ActionOPD with a same-model teacher and privileged context. These results establish ActionOPD as a promising approach to agent distillation.

View arXiv page View PDF Project page GitHub Add to collection

Get this paper in your agent:

hf papers read 2610.19820

Don't have the latest CLI?

curl -LsSf https://hf.co/cli/install.sh | bash

Models citing this paper 0

No model linking this paper

Cite arxiv.org/abs/2610.19820 in a model README.md to link it from this page.

Datasets citing this paper 0

No dataset linking this paper

Cite arxiv.org/abs/2610.19820 in a dataset README.md to link it from this page.

Spaces citing this paper 0

No Space linking this paper

Cite arxiv.org/abs/2610.19820 in a Space README.md to link it from this page.

Collections including this paper 0

No Collection including this paper

Add this paper to a collection to link it from this page.

來源:HuggingFace Daily Papers(社區熱門論文) · huggingface.co