Critique-Guided Distillation for Efficient and Robust Language Model Reasoning

• URL: http://arxiv.org/abs/2505.11628v3
• Date: Fri, 26 Sep 2025 21:35:43 GMT
• Title: Critique-Guided Distillation for Efficient and Robust Language Model Reasoning
• Authors: Berkcan Kapusuzoglu, Supriyo Chakraborty, Chia-Hsuan Lee, Sambit Sahu,
• Abstract summary: Supervised fine-tuning with expert demonstrations often suffers from the imitation problem.<n>We propose Critique-Guided Distillation (CGD), a multi-stage training framework that augments SFT with teacher-generated explanatory critiques and refined responses.<n>Our analyses show that CGD consistently reduces refinement uncertainty, improves alignment between critiques and responses, and enhances sample efficiency.
• Score: 4.8433206430407045
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Supervised fine-tuning (SFT) with expert demonstrations often suffers from the imitation problem, where models reproduce correct responses without internalizing the underlying reasoning. We propose Critique-Guided Distillation (CGD), a multi-stage training framework that augments SFT with teacher-generated explanatory critiques and refined responses. Instead of directly imitating teacher outputs, a student learns to map the triplet of prompt, its own initial response, and teacher critique into the refined teacher response, thereby capturing both what to output and why. Our analyses show that CGD consistently reduces refinement uncertainty, improves alignment between critiques and responses, and enhances sample efficiency. On reasoning benchmarks, CGD achieves substantial gains across LLaMA and Qwen families, including +15.0% on AMC23 and +12.2% on MATH-500, while avoiding the format drift issues observed in prior critique-based fine-tuning. Importantly, on LLaMA-3.1-8B CGD approaches or exceeds the performance of SimpleRL-Zero, which is a DeepSeek-R1 replication, while requiring 60x less compute. Beyond reasoning, CGD maintains or improves general instruction-following and factual accuracy, matching baseline performance on IFEval, MUSR, TruthfulQA, and BBH. In contrast, prior critique-based methods degrade these capabilities (e.g., -21% on IFEval). Taken together, these results establish CGD} as a robust and generalizable alternative to both conventional SFT and RL-based methods, offering a more efficient path toward advancing the reasoning and safety of large language models.

Related papers

• Curriculum Learning for Efficient Chain-of-Thought Distillation via Structure-Aware Masking and GRPO [24.91321958525287]
  Distilling Chain-of-Thought (CoT) reasoning from large language models into compact student models presents a fundamental challenge.<n>Existing approaches either compress reasoning into single-step, losing the interpretability that makes CoT valuable.<n>We present a three-stage curriculum learning framework that addresses this capacity mismatch through progressive skill acquisition.
  arXiv  Detail & Related papers  (2026-02-05T05:27:11Z)
• Teaching Large Reasoning Models Effective Reflection [62.73646680747003]
  Large Reasoning Models (LRMs) have recently shown impressive performance on complex reasoning tasks.<n>However, not all reflections are beneficial-many are superficial, offering little to no improvement over the original answer.<n>We first propose Self-Critique Fine-Tuning (SCFT), a training framework that enhances the model's reflective reasoning ability using only self-generated critiques.
  arXiv  Detail & Related papers  (2026-01-19T04:51:53Z)
• When Actions Teach You to Think: Reasoning-Action Synergy via Reinforcement Learning in Conversational Agents [2.689316553293938]
  Supervised fine-tuning (SFT) has emerged as one of the most effective ways to improve the performance of large language models (LLMs) in downstream tasks.<n>We propose a pipeline in which LLMs generate reasoning steps that guide both the invocation of tools and the final answer generation for conversational agents.
  arXiv  Detail & Related papers  (2025-12-12T04:44:40Z)
• RL from Teacher-Model Refinement: Gradual Imitation Learning for Machine Translation [31.28415780479141]
  Reinforcement Learning from Teacher-Model Refinement (RLfR) is a novel framework that removes reliance on static triplets by leveraging continuous, high-quality feedback from an external teacher model (GPT-4o)<n>On the FLORES-200 benchmark (English to and from German, Spanish, Chinese, Korean, and Japanese), RLfR consistently outperforms both MT-SFT and preference-based baselines.
  arXiv  Detail & Related papers  (2025-07-29T20:35:35Z)
• RefCritic: Training Long Chain-of-Thought Critic Models with Refinement Feedback [57.967762383794806]
  RefCritic is a long-chain-of-thought critic module based on reinforcement learning with dual rule-based rewards.<n>We evaluate RefCritic on Qwen2.5-14B-Instruct and DeepSeek-R1-Distill-Qwen-14B across five benchmarks.
  arXiv  Detail & Related papers  (2025-07-20T16:19:51Z)
• Dual-Expert Consistency Model for Efficient and High-Quality Video Generation [57.33788820909211]
  We propose a parameter-efficient textbfDual-Expert Consistency Model(DCM), where a semantic expert focuses on learning semantic layout and motion, while a detail expert specializes in fine detail refinement.<n>Our approach achieves state-of-the-art visual quality with significantly reduced sampling steps, demonstrating the effectiveness of expert specialization in video diffusion model distillation.
  arXiv  Detail & Related papers  (2025-06-03T17:55:04Z)
• Critique-GRPO: Advancing LLM Reasoning with Natural Language and Numerical Feedback [59.078756231841574]
  Critique-GRPO is an online RL framework that integrates both natural language and numerical feedback for effective policy optimization.<n>We show Critique-GRPO consistently outperforms supervised learning and RL-based fine-tuning methods across eight challenging mathematical, STEM, and general reasoning tasks.
  arXiv  Detail & Related papers  (2025-06-03T17:39:02Z)
• Long-Short Chain-of-Thought Mixture Supervised Fine-Tuning Eliciting Efficient Reasoning in Large Language Models [22.423202755603768]
  We propose textbfLong-textbfShort Chain-of-Thought textbfMixture textbfSupervised textbfFine-textbfTuning.<n>Our experiments demonstrate that models trained using the LS-Mixture SFT method, compared to those trained with direct SFT, achieved an average accuracy improvement of 2.3%.
  arXiv  Detail & Related papers  (2025-05-06T12:18:11Z)
• Retrieval is Not Enough: Enhancing RAG Reasoning through Test-Time Critique and Optimization [58.390885294401066]
  Retrieval-augmented generation (RAG) has become a widely adopted paradigm for enabling knowledge-grounded large language models (LLMs)<n>RAG pipelines often fail to ensure that model reasoning remains consistent with the evidence retrieved, leading to factual inconsistencies or unsupported conclusions.<n>We propose AlignRAG, a novel iterative framework grounded in Critique-Driven Alignment (CDA)<n>We introduce AlignRAG-auto, an autonomous variant that dynamically terminates refinement, removing the need to pre-specify the number of critique iterations.
  arXiv  Detail & Related papers  (2025-04-21T04:56:47Z)
• SEAL: Steerable Reasoning Calibration of Large Language Models for Free [58.931194824519935]
  Large Language Models (LLMs) have demonstrated compelling capabilities for complex reasoning tasks via the extended chain-of-thought (CoT) reasoning mechanism.<n>Recent studies reveal substantial redundancy in the CoT reasoning traces, which negatively impacts model performance.<n>We introduce SEAL, a training-free approach that seamlessly calibrates the CoT process, improving accuracy while demonstrating significant efficiency gains.
  arXiv  Detail & Related papers  (2025-04-07T02:42:07Z)
• TwT: Thinking without Tokens by Habitual Reasoning Distillation with Multi-Teachers' Guidance [42.8895384120507]
  We propose TwT, a method that reduces inference-time costs through habitual reasoning distillation with multi-teachers' guidance.<n>Our approach internalizes explicit reasoning into the model's habitual behavior through a Teacher-Guided compression strategy.<n> Experimental results demonstrate that TwT effectively reduces inference costs while preserving superior performance.
  arXiv  Detail & Related papers  (2025-03-31T15:16:31Z)
• Distill Not Only Data but Also Rewards: Can Smaller Language Models Surpass Larger Ones? [58.80794196076336]
  Distilling large language models (LLMs) typically involves transferring the teacher model's responses through supervised fine-tuning (SFT)<n>We propose a novel distillation pipeline that transfers both responses and rewards.<n>Our method generates pseudo-rewards through a self-supervised mechanism that leverages the inherent structure of both teacher and student responses.
  arXiv  Detail & Related papers  (2025-02-26T20:50:11Z)
• STEP: Enhancing Video-LLMs' Compositional Reasoning by Spatio-Temporal Graph-guided Self-Training [87.58996020705258]
  Video Large Language Models (Video-LLMs) have recently shown strong derivation in basic video understanding tasks.<n>Video-LLMs struggle with compositional reasoning that requires multi-step explicit-temporal inference across object relations, interactions and events.<n>We propose STEP, a novel graph-guided self-training method that enables VideoLLMs to generate reasoning-rich finetuning data from any raw videos to improve itself.
  arXiv  Detail & Related papers  (2024-11-29T11:54:55Z)
• Enhancing LLM Reasoning via Critique Models with Test-Time and Training-Time Supervision [120.40788744292739]
  We propose a two-player paradigm that separates the roles of reasoning and critique models. We first propose AutoMathCritique, an automated and scalable framework for collecting critique data. We demonstrate that the critique models consistently improve the actor's performance on difficult queries at test-time.
  arXiv  Detail & Related papers  (2024-11-25T17:11:54Z)
• Learning to Refine with Fine-Grained Natural Language Feedback [81.70313509881315]
  We propose looking at refinement with feedback as a composition of three distinct LLM competencies. A key property of the proposed Detect, Critique, Refine ("DCR") method is that the step 2 critique model can give fine-grained feedback about errors. We show that models of different capabilities benefit from refining with DCR on the task of improving factual consistency of document grounded summaries.
  arXiv  Detail & Related papers  (2024-07-02T16:15:01Z)
• A Critical Evaluation of AI Feedback for Aligning Large Language Models [60.42291111149438]
  We show that simple supervised fine-tuning with GPT-4 as the teacher outperforms existing RLAIF pipelines. More generally, we find that the gains from RLAIF vary substantially across base model families, test-time evaluation protocols, and critic models.
  arXiv  Detail & Related papers  (2024-02-19T18:53:54Z)
• Training Chain-of-Thought via Latent-Variable Inference [30.21067593018967]
  Large language models (LLMs) solve problems more accurately and interpretably when instructed to work out the answer step by step using a chain-of-thought'' prompt. Naively combining CoT with supervised tuning requires supervision not just of the correct answers, but also of detailed rationales that lead to those answers. We propose a fine-tuning strategy that tries to maximize the emphmarginal log-likelihood of generating a correct answer using CoT prompting.
  arXiv  Detail & Related papers  (2023-11-28T17:47:32Z)
• Constructive Large Language Models Alignment with Diverse Feedback [76.9578950893839]
  We introduce Constructive and Diverse Feedback (CDF) as a novel method to enhance large language models alignment. We exploit critique feedback for easy problems, refinement feedback for medium problems, and preference feedback for hard problems. By training our model with this diversified feedback, we achieve enhanced alignment performance while using less training data.
  arXiv  Detail & Related papers  (2023-10-10T09:20:14Z)
• OPT-R: Exploring the Role of Explanations in Finetuning and Prompting for Reasoning Skills of Large Language Models [48.412284346337344]
  We conduct a thorough investigation into the reasoning capabilities of Large Language Models (LLMs) Our study entails finetuning three different sizes of Open Pretrained Transformers (OPT) We then evaluate all models on 57 out-of-domain tasks drawn from the SUPER-NATURALINSTRUCTIONS benchmark.
  arXiv  Detail & Related papers  (2023-05-19T20:58:22Z)
• Referee: Reference-Free Sentence Summarization with Sharper Controllability through Symbolic Knowledge Distillation [72.70058049274664]
  We present Referee, a novel framework for sentence summarization that can be trained reference-free (i.e., requiring no gold summaries for supervision) Our work is the first to demonstrate that reference-free, controlled sentence summarization is feasible via the conceptual framework of Symbolic Knowledge Distillation.
  arXiv  Detail & Related papers  (2022-10-25T07:07:54Z)
• Noisy Self-Knowledge Distillation for Text Summarization [83.49809205891496]
  We apply self-knowledge distillation to text summarization which we argue can alleviate problems with maximum-likelihood training. Our student summarization model is trained with guidance from a teacher which generates smoothed labels to help regularize training. We demonstrate experimentally on three benchmarks that our framework boosts the performance of both pretrained and non-pretrained summarizers.
  arXiv  Detail & Related papers  (2020-09-15T12:53:09Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.