MiCoTA: Bridging the Learnability Gap with Intermediate CoT and Teacher Assistants

• URL: http://arxiv.org/abs/2507.01887v1
• Date: Wed, 02 Jul 2025 16:57:01 GMT
• Title: MiCoTA: Bridging the Learnability Gap with Intermediate CoT and Teacher Assistants
• Authors: Dongyi Ding, Tiannan Wang, Chenghao Zhu, Meiling Tao, Yuchen Eleanor Jiang, Wangchunshu Zhou,
• Abstract summary: We introduce textbfMid-textbfCoT textbfTeacher textbfAssistant Distillation (MiCoTAl)<n>MiCoTAl is a framework for improving long CoT distillation for small language models (SLMs)
• Score: 25.45861816665351
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Large language models (LLMs) excel at reasoning tasks requiring long thought sequences for planning, reflection, and refinement. However, their substantial model size and high computational demands are impractical for widespread deployment. Yet, small language models (SLMs) often struggle to learn long-form CoT reasoning due to their limited capacity, a phenomenon we refer to as the "SLMs Learnability Gap". To address this, we introduce \textbf{Mi}d-\textbf{Co}T \textbf{T}eacher \textbf{A}ssistant Distillation (MiCoTAl), a framework for improving long CoT distillation for SLMs. MiCoTA employs intermediate-sized models as teacher assistants and utilizes intermediate-length CoT sequences to bridge both the capacity and reasoning length gaps. Our experiments on downstream tasks demonstrate that although SLMs distilled from large teachers can perform poorly, by applying MiCoTA, they achieve significant improvements in reasoning performance. Specifically, Qwen2.5-7B-Instruct and Qwen2.5-3B-Instruct achieve an improvement of 3.47 and 3.93 respectively on average score on AIME2024, AMC, Olympiad, MATH-500 and GSM8K benchmarks. To better understand the mechanism behind MiCoTA, we perform a quantitative experiment demonstrating that our method produces data more closely aligned with base SLM distributions. Our insights pave the way for future research into long-CoT data distillation for SLMs.

Related papers

• TL;DR: Too Long, Do Re-weighting for Efficient LLM Reasoning Compression [55.37723860832064]
  We propose a dynamic ratio-based training pipeline that does not rely on sophisticated data annotations.<n>We validate our approach across models on DeepSeek-R1-Distill-7B and DeepSeek-R1-Distill-14B and on a diverse set of benchmarks with varying difficulty levels.
  arXiv  Detail & Related papers  (2025-06-03T09:23:41Z)
• Distilling LLM Agent into Small Models with Retrieval and Code Tools [57.61747522001781]
  Agent Distillation is a framework for transferring reasoning capability and task-solving behavior from large language models into small language models.<n>Our results show that sLMs as small as 0.5B, 1.5B, 3B parameters can achieve performance competitive with next-tier larger 1.5B, 3B, 7B models.
  arXiv  Detail & Related papers  (2025-05-23T08:20:15Z)
• Phi-4-Mini-Reasoning: Exploring the Limits of Small Reasoning Language Models in Math [135.1260782461186]
  Chain-of-Thought (CoT) significantly enhances formal reasoning capabilities in Large Language Models (LLMs)<n>However, improving reasoning in Small Language Models (SLMs) remains challenging due to their limited model capacity.<n>We present a systematic training recipe for SLMs that consists of four steps: (1) large-scale mid-training on diverse distilled long-CoT data, (2) supervised fine-tuning on high-quality long-CoT data, (3) Rollout DPO leveraging a carefully curated preference dataset, and (4) Reinforcement Learning (RL) with Verifiable Reward.
  arXiv  Detail & Related papers  (2025-04-30T00:04:35Z)
• MA-LoT: Model-Collaboration Lean-based Long Chain-of-Thought Reasoning enhances Formal Theorem Proving [30.112351299773632]
  We propose a comprehensive framework for Lean4 theorem proving to solve this issue.<n>It separates the cognition tasks of general NL for whole-proof generation and error analysis for proof correction.<n>Our framework achieves a **61.07%** accuracy rate on the Lean4 version of the MiniF2F-Test dataset.
  arXiv  Detail & Related papers  (2025-03-05T05:50:31Z)
• Marco-o1 v2: Towards Widening The Distillation Bottleneck for Reasoning Models [39.22557129190619]
  Distillation--post-training on LRMs-generated data--is a straightforward yet effective method to enhance the reasoning abilities of smaller models.<n>To alleviate this bottleneck, we propose constructing tree-based CoT data from scratch via Monte Carlo Tree Search.<n>We then exploit a set of CoT-aware approaches, including Thoughts Length Balance, Fine-grained DPO, and Joint Post-training Objective, to enhance SFT and RL on the constructed data.
  arXiv  Detail & Related papers  (2025-03-03T12:17:36Z)
• Unveiling the Key Factors for Distilling Chain-of-Thought Reasoning [33.02060729778806]
  This study examines the factors influencing Chain-of-Thought (CoT) distillation in Small Language Models (SLMs)<n>We find that SLMs exhibit a non-monotonic relationship with granularity, with stronger models benefiting from finer-grained reasoning and weaker models performing better with simpler CoT supervision.<n>These findings emphasize the need to tailor CoT strategies to specific student model, offering actionable insights for optimizing CoT distillation in SLMs.
  arXiv  Detail & Related papers  (2025-02-25T09:08:45Z)
• LLMs Can Easily Learn to Reason from Demonstrations Structure, not content, is what matters! [53.84130385074551]
  Large reasoning models (LRMs) tackle complex reasoning problems by following long chain-of-thoughts (Long CoT)<n>We find that a Large Language model (LLM) can effectively learn Long CoT reasoning through data-efficient supervised fine-tuning (SFT) and parameter-efficient low-rank adaptation (LoRA)<n>With just 17k long CoT training samples, the Qwen2.5-32B-Instruct model achieves significant improvements on a wide range of math and coding benchmarks.
  arXiv  Detail & Related papers  (2025-02-11T08:48:48Z)
• BOLT: Bootstrap Long Chain-of-Thought in Language Models without Distillation [88.77999917897702]
  o1 from OpenAI has demonstrated remarkable reasoning capabilities.<n>Many teams have attempted to replicate its LongCoT and reasoning capabilities.<n>This paper introduces a novel approach to enable LLM's LongCoT capacity without distillation from o1-like models or expensive human annotations.
  arXiv  Detail & Related papers  (2025-02-06T08:19:59Z)
• Understanding Chain-of-Thought in LLMs through Information Theory [16.78730663293352]
  We formalize Chain-of-Thought (CoT) reasoning in Large Language Models (LLMs) through an information-theoretic lens.<n>Specifically, our framework quantifies the information-gain' at each reasoning step, enabling the identification of failure modes.<n>We demonstrate the efficacy of our approach through extensive experiments on toy arithmetic, GSM8K and PRM800k datasets.
  arXiv  Detail & Related papers  (2024-11-18T19:14:36Z)
• MindStar: Enhancing Math Reasoning in Pre-trained LLMs at Inference Time [51.5039731721706]
  MindStar is a purely inference-based searching method for large language models. It formulates reasoning tasks as searching problems and proposes two search ideas to identify the optimal reasoning paths. It significantly enhances the reasoning abilities of open-source models, such as Llama-2-13B and Mistral-7B, and achieves comparable performance to GPT-3.5 and Grok-1.
  arXiv  Detail & Related papers  (2024-05-25T15:07:33Z)
• Mixed Distillation Helps Smaller Language Model Better Reasoning [27.934081882868902]
  We introduce Mixed Distillation (MD) framework, which capitalizes on the strengths of Program of Thought (PoT) and Chain of Thought (CoT) capabilities within large language models (LLMs) Our experimental results show that MD significantly enhances the single-path and multi-path reasoning ability of smaller models in various tasks.
  arXiv  Detail & Related papers  (2023-12-17T14:28:28Z)
• Multistage Collaborative Knowledge Distillation from a Large Language Model for Semi-Supervised Sequence Generation [60.657065293413716]
  We study semi-supervised sequence generation tasks, where the few labeled examples are too scarce to finetune a model. We present the discovery that a student model distilled from a few-shot prompted large language models can commonly generalize better than its teacher.
  arXiv  Detail & Related papers  (2023-11-15T01:28:28Z)

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.