# Think When You Need: Self-Adaptive Chain-of-Thought Learning

arXiv: https://arxiv.org/abs/2504.03234 · HTML: https://mlsanity.com/paper/2504.03234/

## Key facts

- Cited by 38 papers in the map (data through 14 Oct 2026).
- Placed in Prompting and chain-of-thought reasoning, one of 17 directions in LLM reasoning.
- Builds on 12 of its 16 references that are in the map.
- First posted on arXiv 4 Apr 2025.

## Placed in

- LLM reasoning / [Prompting and chain-of-thought reasoning](https://mlsanity.com/lr/prompting-and-chain-of-thought-reasoning/)

## Builds on (most cited first, 12 of 12)

- [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://mlsanity.com/paper/2402.03300/) (2024)
- [DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning](https://mlsanity.com/paper/2501.12948/) (2025)
- [Qwen2.5 Technical Report](https://mlsanity.com/paper/2412.15115/) (2024)
- [DAPO: An Open-Source LLM Reinforcement Learning System at Scale](https://mlsanity.com/paper/2503.14476/) (2025)
- [HybridFlow: A Flexible and Efficient RLHF Framework](https://mlsanity.com/paper/2409.19256/) (2024)
- [OpenAI o1 System Card](https://mlsanity.com/paper/2412.16720/) (2024)
- [Kimi k1.5: Scaling Reinforcement Learning with LLMs](https://mlsanity.com/paper/2501.12599/) (2025)
- [LLM-Blender: Ensembling Large Language Models with Pairwise Ranking and Generative Fusion](https://mlsanity.com/paper/2306.02561/) (2023)
- [L1: Controlling How Long A Reasoning Model Thinks With Reinforcement Learning](https://mlsanity.com/paper/2503.04697/) (2025)
- [O1-Pruner: Length-Harmonizing Fine-Tuning for O1-Like Reasoning Pruning](https://mlsanity.com/paper/2501.12570/) (2025)
- [PairJudge RM: Perform Best-of-N Sampling with Knockout Tournament](https://mlsanity.com/paper/2501.13007/) (2025)
- [Adaptive Group Policy Optimization: Towards Stable Training and Token-Efficient Reasoning](https://mlsanity.com/paper/2503.15952/) (2025)

## Built on by (most cited first, 30 of 38)

- [Efficient Reasoning Models: A Survey](https://mlsanity.com/paper/2504.10903/) (2025)
- [R-KV: Redundancy-aware KV Cache Compression for Reasoning Models](https://mlsanity.com/paper/2505.24133/) (2025)
- [Don't Overthink It: A Survey of Efficient R1-style Large Reasoning Models](https://mlsanity.com/paper/2508.02120/) (2025)
- [R1-Compress: Long Chain-of-Thought Compression via Chunk Compression and Search](https://mlsanity.com/paper/2505.16838/) (2025)
- [A Comprehensive Survey on Trustworthiness in Reasoning with Large Language Models](https://mlsanity.com/paper/2509.03871/) (2025)
- [VeriThinker: Learning to Verify Makes Reasoning Model Efficient](https://mlsanity.com/paper/2505.17941/) (2025)
- [Towards Concise and Adaptive Thinking in Large Reasoning Models: A Survey](https://mlsanity.com/paper/2507.09662/) (2025)
- [MUR: Momentum Uncertainty guided Reasoning for Large Language Models](https://mlsanity.com/paper/2507.14958/) (2025)
- [Do Thinking Tokens Help or Trap? Towards More Efficient Large Reasoning Model](https://mlsanity.com/paper/2506.23840/) (2025)
- [Reconsidering Overthinking: Penalizing Internal and External Redundancy in CoT Reasoning](https://mlsanity.com/paper/2508.02178/) (2025)
- [Don't Think Longer, Think Wisely: Optimizing Thinking Dynamics for Large Reasoning Models](https://mlsanity.com/paper/2505.21765/) (2025)
- [LAPO: Internalizing Reasoning Efficiency via Length-Adaptive Policy Optimization](https://mlsanity.com/paper/2507.15758/) (2025)
- [Exploring the System 1 Thinking Capability of Large Reasoning Models](https://mlsanity.com/paper/2504.10368/) (2025)
- [Understanding and Steering the Cognitive Behaviors of Reasoning Models at Test-Time](https://mlsanity.com/paper/2512.24574/) (2025)
- [How Far Are We from Optimal Reasoning Efficiency?](https://mlsanity.com/paper/2506.07104/) (2025)
- [Probing the Difficulty Perception Mechanism of Large Language Models](https://mlsanity.com/paper/2510.05969/) (2025)
- [Neural Chain-of-Thought Search: Searching the Optimal Reasoning Path to Enhance Large Language Models](https://mlsanity.com/paper/2601.11340/) (2026)
- [VTC-R1: Vision-Text Compression for Efficient Long-Context Reasoning](https://mlsanity.com/paper/2601.22069/) (2026)
- [Efficient Long CoT Reasoning in Small Language Models](https://mlsanity.com/paper/2505.18440/) (2025)
- [Enhancing Large Language Models through Structured Reasoning](https://mlsanity.com/paper/2506.20241/) (2025)
- [TokenSqueeze: Performance-Preserving Compression for Reasoning LLMs](https://mlsanity.com/paper/2511.13223/) (2025)
- [Correct, Concise and Complete: Multi-stage Training For Adaptive Reasoning](https://mlsanity.com/paper/2601.02972/) (2026)
- [Efficient Reasoning with Balanced Thinking](https://mlsanity.com/paper/2603.12372/) (2026)
- [Verifying Meta-Awareness via Predictive Rewards in Reasoning Models](https://mlsanity.com/paper/2510.03259/) (2025)
- [ESTAR: Early-Stopping Token-Aware Reasoning For Efficient Inference](https://mlsanity.com/paper/2602.10004/) (2026)
- [Adaptive Dual Reasoner: Large Reasoning Models Can Think Efficiently by Hybrid Reasoning](https://mlsanity.com/paper/2510.10207/) (2025)
- [Compress-Distill: Reasoning Trace Compression for Efficient Knowledge Distillation](https://mlsanity.com/paper/2606.05988/) (2026)
- [SCALE: Selective Resource Allocation for Overcoming Performance Bottlenecks in Mathematical Test-time Scaling](https://mlsanity.com/paper/2512.00466/) (2025)
- [RANKVIDEO: Reasoning Reranking for Text-to-Video Retrieval](https://mlsanity.com/paper/2602.02444/) (2026)
- [Towards Generalizable Reasoning: Group Causal Counterfactual Policy Optimization for LLM Reasoning](https://mlsanity.com/paper/2602.06475/) (2026)

## Abstract (arXiv)

Chain of Thought (CoT) reasoning enhances language models' performance but often leads to inefficient "overthinking" on simple problems. We identify that existing approaches directly penalizing reasoning length fail to account for varying problem complexity. Our approach constructs rewards through length and quality comparisons, guided by theoretical assumptions that jointly enhance solution correctness with conciseness. Moreover, we further demonstrate our method to fuzzy tasks where ground truth is unavailable. Experiments across multiple reasoning benchmarks demonstrate that our method maintains accuracy while generating significantly more concise explanations, effectively teaching models to "think when needed."