Reasoning Distillation

Follow this topic Ask about this topic

Reasoning distillation trains a smaller student to reproduce or benefit from a larger model’s problem-solving, using traces, token probabilities, feedback, or reward signals as supervision. Common approaches include supervised trace imitation, on-policy distillation, reinforcement learning, and curricula that adapt examples or teacher signals to the student’s capabilities. The goal is to retain useful reasoning in settings such as mathematics, coding, multimodal tasks, and safety while reducing training or inference costs. A central challenge is that copying a teacher’s behavior can pass along flawed steps or overwhelm a student, and gains in solving do not always translate into reliable confidence, answer formatting, or stopping.

Latest papers

23 papers3 papers in the last four weeks, down 25% on the four weeks before. 0.0% of all new papers.

From Solver Trajectories to Teaching Trajectories: Cognition-Aligned Reasoning Distillation
Oct 27, 2026Yashuo Luo, Qianren Mao, Ziqi Qin +6Reasoning Distillation Language Model Distillation
National College for Excellent Engineers, Beihang University · Zhongguancun Laboratory · Beihang University +3
Bridged Self-Distillation for On-Policy Learning
Oct 19, 2026Runsheng Chen, Yihong Liu, Raoyuan Zhao +3Reasoning Distillation On-Policy Distillation
Technical University of Munich, Germany · Technical University of Munich · LMU Munich, Germany +3
Collaborative Reasoning Distillation via Cross-Feedback and Coherent Curation
Oct 7, 2026Taehoon Kim, Seunggeun Cho, Dongsu HanReasoning Distillation LLM Reasoning
KAIST
Solving Without Stopping: On-Policy Distillation at Small Scale
Sep 29, 2026Hongyang Li, Yiming Zhu, Xiao Li +3Reasoning Distillation Small Language Models
University of Luxembourg · Seafill Open-Source Community · Universit´e Paris-Saclay
ChemOPD: Multi-Teacher On-Policy Distillation for Multi-Task Chemical Reasoning
Sep 27, 2026Yaoyao Xu, Xinjian Zhao, Xiaozhuang Song +2Chemical Reasoning in LLMs Reasoning Distillation
School of Data Science, The Chinese University of Hong Kong, Shenzhen · Shanghai Artificial Intelligence Laboratory
TeacherGRPO: Closing the Capacity Gap in Reasoning Distillation via Teacher Alignment
Sep 27, 2026Zhenyu Lei, Zihan Chen, Yaochen Zhu +5Reasoning Distillation RL for Language Model Reasoning
University of Virginia · Netflix · University of Washington +2
On Repulsive and Attractive Teachers: Separating Correctness from Behavior in Self-Distillation
Sep 18, 2026Anton Baumann, Akmal Ashirmatov, Leo Schmidt-Traub +4Reasoning Distillation On-Policy Self-Distillation
ETH Zurich · Max Planck Institute for Intelligent Systems
Distill Globally, Adapt Locally: Reasoning Distillation and Product-Type Test-Time Training for Scalable Trade-Up Recommendation
Sep 4, 2026Siliang Liu, Mohammad Ghasemi, Sapan Patel +1Reasoning Distillation Large Language Model-Based Recommendation
Amazon Everyday Essentials Technologies
Reason in the Words You Speak: Idiolectal Paraphrasing Off-Policy Traces for Reasoning Distillation in VideoLLMs
Aug 27, 2026Ji Soo Lee, Jinyoung Park, Seohyun Lee +4Reasoning Distillation VLM Reasoning
KAIST · Korea University · Google Cloud AI Research
Benchmarking and Reasoning Distillation of Large Language Models for Feedback Controller Design in Complex Dynamical Systems
Aug 7, 2026Zhongchao Zhou, Yixuan Xie, Wenwei Yu +5LLM Evaluation Language Model-Based Control
Matsuo-Iwasawa Laboratory, Department of Technology Management for Innovation, School of Engineering, The University of Tokyo, Tokyo, Japan · Department of Statistics, University of Illinois Urbana-Champaign, Champaign, IL, USA · Center for Frontier Medical Engineering, Chiba University, Chiba, Japan +1
Distilling Reasoning Traces into Advisory Prompts for Software Engineering Tasks
Aug 1, 2026Faizan Faisal, Prem Devanbu, Toufique AhmedReasoning Distillation LLM Prompting
University of California, Davis · IBM
DT-Guard: Intent-Driven Reasoning-Active Training for Reasoning-Free LLM Safety Guardrail
Jul 7, 2026He Liu, Changtao Miao, Xinjie Yang +12Reasoning Distillation LLM Guardrails
Ant Digital Technologies, Ant Group
Z-Reward: Beyond Scalar Rewards by Internalizing Reasoning into Score Distributions
Jun 8, 2026Xin Jin, Huanqia Cai, Zhen Li +9Reward Modeling Reasoning Distillation
Z-Image Team, Alibaba Group · VCIP, CS, Nankai University
Tailoring the Curriculum: Student-Centered Reasoning Distillation via Dynamic Data-Model Compatibility
May 28, 2026Jiahao Huang, Fei Cheng, Junfeng Jiang +1Reasoning Distillation Training Data Selection
University of Tokyo · Kyoto University · National Institute of Informatics
Decoupling KL and Trajectories: A Unified Perspective for SFT, DAgger, Offline RL, and OPD in LLM Distillation
May 16, 2026Anhao Zhao, Haoran Xin, Yingqi Fan +3Reasoning Distillation On-Policy Distillation
Eastern Institute of Technology, Ningbo · Thrust of Artificial Intelligence, The Hong Kong University of Science and Technology (Guangzhou) · Shanghai Jiao Tong University +1
Curriculum Learning-Guided Progressive Distillation in Large Language Models
May 11, 2026Jincheng Cao, Fanzhi Zeng, Leqi Liu +1Reasoning Distillation Language Model Distillation
The University of Texas at Austin · Google Research
ReflectMT: Internalizing Reflection for Efficient and High-Quality Machine Translation
Apr 21, 2026Kunquan Li, Yingxue Zhang, Fandong Meng +1Reasoning Distillation Machine Translation
School of Informatics, Xiamen University, China · WeChat AI, Tencent Inc, China
Validity-Calibrated Reasoning Distillation
Apr 14, 2026Khouloud Saadi, Di WangReasoning Distillation Language Model Distillation
Department of Computer Science, KAUST, Thuwal, Saudi Arabia
Revisiting the Capacity Gap in Chain-of-Thought Distillation from a Practical Perspective
Apr 10, 2026Tokio Kajitsuka, Ukyo Honda, Sho TakaseCoT Distillation Reasoning Distillation
The University of Tokyo, Tokyo, Japan · CyberAgent, Tokyo, Japan
Less Data Approximates More: Earning Faithful Confidence in High-Stakes Domains
Apr 9, 2026Haokai Ma, Lee Yan Zhen, Gang Yang +4RL for Language Models Reinforcement Learning
National University of Singapore, Singapore · Singapore Management University, Singapore
LLM-MRD: LLM-Guided Multi-View Reasoning Distillation for Fake News Detection
Mar 10, 2026Weilin Zhou, Shanwen Tan, Enhao Gu +1Fake News Detection Reasoning Distillation
Xinjiang University, Urumqi, China · Joint International Research Laboratory of Silk Road Multilingual Cognitive Computing, Xinjiang University, Urumqi, China · Sichuan University, Chengdu, China
Probing to Refine: Reinforcement Distillation of LLMs via Explanatory Inversion
Feb 26, 2026Zhen Tan, Chengshuai Zhao, Song Wang +3Language Model Distillation Reasoning Distillation
Arizona State University · University of Virginia · University of North Carolina at Chapel Hill
Federated Reasoning Distillation Framework with Model Learnability-Aware Data Allocation
Feb 21, 2026Wei Guo, Siyuan Lu, Xiangdong Ran +8Reasoning Distillation Federated Knowledge Distillation
Beihang University, Beijing 100191, China · Heilongjiang University, Harbin 150080, China · Beijing University of Civil Engineering and Architecture, Beijing 100044, China +3

Oct 27, 2026J/K move · Enter open · S save

From Solver Trajectories to Teaching Trajectories: Cognition-Aligned Reasoning Distillation

Yashuo Luo, Qianren Mao, Ziqi Qin, Junnan Liu+5

National College for Excellent Engineers, Beihang University · Zhongguancun Laboratory · Beihang University+3

Reasoning Distillation Language Model Distillation CoT Distillation RL for Knowledge Distillation

Open paper AskPDF arXiv

Knowledge distillation from large reasoning models (LRMs) often uses raw teacher trajectories as supervision. However, this practice creates a mismatch between the goal of distillation and the nature of teacher trajectories. Advanced LRMs typically improve their reasoning capabilities through reward-driven post-training, such as RLHF. This training optimizes them as strong solvers, rather than as teachers that produce reasoning processes smaller models can easily learn. Since smaller models have different "cognitive capacities" compared to their larger teachers, directly imitating raw teacher trajectories can sometimes be ineffective and often requires a substantial number of high-quality samples. To bridge this mismatch, we propose Cognition-Aligned Reasoning Distillation (CARD), an RL-based framework that transforms solver trajectories into teaching trajectories aligned with the student's cognition. Avoiding expensive resampling, our framework directly transforms existing trajectories. We optimize CARD with two reward signals: (1) a Distribution Alignment Reward that matches the student's predictive distribution, and (2) a Logical Coherence Reward that reduces misleading failed reasoning steps. Comprehensive evaluations across multiple student models and settings demonstrate that CARD outperforms distillation from raw teacher trajectories in in-domain settings while achieving substantial gains in out-of-distribution performance. Code is available at https://github.com/LiteSSLHub/CARD.