CtrlCoT: Dual-Granularity Chain-of-Thought Compression for Controllable Reasoning

2026-01-28 · Zhenxuan Fan, Jie Cao, Yang Dai, Zheqi Lv, Wenqiao Zhang, Zhongle Xie, Peng LU, Beng Chin Ooi arxiv

Chain-of-thought (CoT) prompting improves LLM reasoning but incurs high latency and memory cost due to verbose traces, motivating CoT compression with preserved correctness. Existing methods either shorten CoTs at the semantic level, which is often conservative, or prune tokens aggressively, which can miss task-critical cues and degrade accuracy. Moreover, combining the two is non-trivial due to sequential dependency, task-agnostic pruning, and distribution mismatch. We propose \textbf{CtrlCoT}, a dual-granularity CoT compression framework that harmonizes semantic abstraction and token-level pruning through three components: Hierarchical Reasoning Abstraction produces CoTs at multiple semantic granularities; Logic-Preserving Distillation trains a logic-aware pruner to retain indispensable reasoning cues (e.g., numbers and operators) across pruning ratios; and Distribution-Alignment Generation aligns compressed traces with fluent inference-time reasoning styles to avoid fragmentation. On MATH-500 with Qwen2.5-7B-Instruct, CtrlCoT uses 30.7\% fewer tokens while achieving 7.6 percentage points higher than the strongest baseline, demonstrating more efficient and reliable reasoning. Our code will be publicly available at https://github.com/fanzhenxuan/Ctrl-CoT.

📄 PDF Abstract BibTeX arXiv:2601.20467

Code (0)

등록된 구현이 없습니다.

When Compression Helps and When It Hurts: Condition-Aware Analysis of Chain-of-Thought Distillation

2026-06-19 · Siyang Lyu, Zhijing Sun, Xinghao Chen, Tong Liu 외 arxiv

Chain-of-Thought (CoT) distillation transfers multi-step reasoning from large reasoning models to smaller students, but verbose teacher traces inflate both training and inference cost. Existing CoT compression methods fa…

HiVid-Narrator: Hierarchical Video Narrative Generation with Scene-Primed ASR-anchored Compression

2026-01-12 · Haoxuan Li, Mengyan Li, Junjun Zheng arxiv

Generating structured narrations for real-world e-commerce videos requires models to perceive fine-grained visual details and organize them into coherent, high-level stories--capabilities that existing approaches struggl…

Video Captioning

Accordion-Thinking: Self-Regulated Step Summaries for Efficient and Readable LLM Reasoning

2026-02-03 · Zhicheng Yang, Zhijiang Guo, Yinya Huang, Yongxin Wang 외 arxiv

Scaling test-time compute via long Chain-of-Thought unlocks remarkable gains in reasoning capabilities, yet it faces practical limits due to the linear growth of KV cache and quadratic attention complexity. In this paper…

Reinforcement Learning

Chain-of-Thought Compression Should Not Be Blind: V-Skip for Efficient Multimodal Reasoning via Dual-Path Anchoring

2026-01-20 · Dongxu Zhang, Yiding Sun, Cheng Tan, Wenbiao Yan 외 arxiv

While Chain-of-Thought (CoT) reasoning significantly enhances the performance of Multimodal Large Language Models (MLLMs), its autoregressive nature incurs prohibitive latency constraints. Current efforts to mitigate thi…

Multimodal Reasoning

From Long to Lean: Performance-aware and Adaptive Chain-of-Thought Compression via Multi-round Refinement

2025-09-26 · Jianzhi Yan, Le Liu, Youcheng Pan, Shiwei Chen 외 arxiv

Chain-of-Thought (CoT) reasoning improves performance on complex tasks but introduces significant inference latency due to verbosity. We propose Multiround Adaptive Chain-of-Thought Compression (MACC), a framework that l…