mlsanityData to 14 Oct 2026

Entropy-based Exploration Conduction for Multi-step Reasoning

arXiv 2503.15848 ↗First posted 20 Mar 2025cs.AI cs.CLcited by 29

On the map: LLM reasoning → Test-time compute and efficient thinking

Abstract (arXiv)

Multi-step processes via large language models (LLMs) have proven effective for solving complex reasoning tasks. However, the depth of exploration of the reasoning procedure can significantly affect the task performance. Existing methods to automatically decide the depth often lead to high cost and a lack of flexibility. To address these issues, we propose Entropy-based Exploration Depth Conduction (Entro-duction), a novel method that dynamically adjusts the exploration depth during multi-step reasoning by monitoring LLM's output entropy and variance entropy. We employ these two features to capture the model's uncertainty of the current step and the fluctuation of uncertainty across consecutive reasoning steps. Based on the observed entropy changes, the LLM selects whether to deepen, expand, or stop exploration according to the probability, which facilitates the trade-off between the reasoning accuracy and exploration effectiveness. Experimental results across four benchmark datasets demonstrate the efficacy of Entro-duction.

Key facts. Cited by 29 papers in the map (data through 14 Oct 2026). Placed in Test-time compute and efficient thinking, one of 17 directions in LLM reasoning. Builds on 36 of its 48 references that are in the map. First posted on arXiv 20 Mar 2025.

Builds on

36 of 48 references.

Closest ancestors · 7

Same direction, newest first

  1. 2025CoT-Valve: Length-Compressible Chain-of-Thought Tuning13 Feb
  2. 2025Demystifying Long Chain-of-Thought Reasoning in LLMs5 Feb
  3. 2024Beyond Introspection: Reinforcing Thinking via Externalist Behavioral Feedback31 Dec
  4. 2024Internal Consistency and Self-Feedback in Large Language Models: A Survey19 Jul
  5. 2024On the Self-Verification Limitations of Large Language Models on Reasoning and Planning Tasks12 Feb
  6. 2023Tree of Uncertain Thoughts Reasoning for Large Language Models14 Sep
  7. 2023Automatically Correcting Large Language Models: Surveying the landscape of diverse self-correction strategies6 Aug

Other references · 29

  1. 2020Language Models are Few-Shot Learnerscited by 32,419
  2. 2023LLaMA: Open and Efficient Foundation Language Modelscited by 14,132
  3. 2021Training Verifiers to Solve Math Word Problemscited by 10,109
  4. 2022Self-Consistency Improves Chain of Thought Reasoning in Language Modelscited by 5,708
  5. 2023Let's Verify Step by Stepcited by 3,789
  6. 2018CommonsenseQA: A Question Answering Challenge Targeting Commonsense Knowledgecited by 1,822
  7. 2022Least-to-Most Prompting Enables Complex Reasoning in Large Language Modelscited by 1,407
  8. 2021Are NLP Models really able to Solve Simple Math Word Problems?cited by 878
  9. 2021Did Aristotle Use a Laptop? A Question Answering Benchmark with Implicit Reasoning Strategiescited by 743
  10. 2022Complexity-Based Prompting for Multi-Step Reasoningcited by 388
  11. 2024To CoT or not to CoT? Chain-of-thought helps mainly on math and symbolic reasoningcited by 263
  12. 2024Chain of Thought Empowers Transformers to Solve Inherently Serial Problemscited by 260
17 more
  1. 2023Specializing Smaller Language Models towards Multi-Step Reasoningcited by 251
  2. 2024Chain-of-Thought Reasoning Without Promptingcited by 210
  3. 2022The Carbon Footprint of Machine Learning Training Will Plateau, Then Shrinkcited by 184
  4. 2020Unsupervised Commonsense Question Answering with Self-Talkcited by 161
  5. 2024Distilling System 2 into System 1cited by 159
  6. 2024Multi-Step Reasoning with Large Language Models, a Surveycited by 143
  7. 2024Chain of Thoughtlessness? An Analysis of CoT in Planningcited by 103
  8. 2023GPT-4 Doesn't Know It's Wrong: An Analysis of Iterative Prompting for Reasoning Problemscited by 89
  9. 2023ConvGQR: Generative Query Reformulation for Conversational Searchcited by 59
  10. 2024Can Language Models Learn to Skip Steps?cited by 59
  11. 2024Beyond Chain-of-Thought: A Survey of Chain-of-X Paradigms for LLMscited by 48
  12. 2024Learning to Trust Your Feelings: Leveraging Self-awareness in LLMs for Hallucination Mitigationcited by 44
  13. 2024Understanding Chain-of-Thought in LLMs through Information Theorycited by 37
  14. 2024Evolutionary Large Language Model for Automated Feature Transformationcited by 29
  15. 2024Prototypical Reward Network for Data-Efficient RLHFcited by 22
  16. 2024Revolutionizing Biomarker Discovery: Leveraging Generative AI for Bio-Knowledge-Embedded Continuous Space Explorationcited by 11
  17. 2024ZEBRA: Zero-Shot Example-Based Retrieval Augmentation for Commonsense Question Answeringcited by 3

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 9; 2026: 20.

29 papers.

Sort:

By area:

  1. 2025SRPO: Enhancing Multimodal LLM Reasoning via Reflection-Aware Reinforcement Learningcited by 57
  2. 2025A Survey of Slow Thinking-based Reasoning LLMs using Reinforced Learning and Inference-time Scaling Lawcited by 19
  3. 2026GlimpRouter: Efficient Collaborative Inference by Glimpsing One Token of Thoughtscited by 17
  4. 2025REVISOR: Beyond Textual Reflection, Towards Multimodal Introspective Reasoning in Long-Form Video Understandingcited by 16
  5. 2025UniConv: Unifying Retrieval and Response Generation for Large Language Models in Conversationscited by 12
  6. 2026OpenDecoder: Open Large Language Model Decoding to Incorporate Document Quality in RAGcited by 11
  7. 2026Reward Modeling for Reinforcement Learning-Based LLM Reasoning: Design, Challenges, and Evaluationcited by 7
  8. 2026Revisiting Text Ranking in Deep Researchcited by 7
  9. 2025Fast Quiet-STaR: Thinking Without Thought Tokenscited by 5
  10. 2025RAST: Reasoning Activation in LLMs via Small-model Transfercited by 5
  11. 2025Dual-Uncertainty Guided Policy Learning for Multimodal Reasoningcited by 5
  12. 2026ReAttn: Improving Attention-based Re-ranking via Attention Re-weightingcited by 5
17 more
  1. 2026StaRPO: Stability-Augmented Reinforcement Policy Optimizationcited by 5
  2. 2026Efficient Reasoning with Balanced Thinkingcited by 4
  3. 2026Unveiling the Entropy Dynamics of Chain-of-Thought Reasoningcited by 4
  4. 2026The Stepwise Informativeness Assumption: Why are Entropy Dynamics and Reasoning Correlated in LLMs?cited by 3
  5. 2025SoT: Structured-of-Thought Prompting Guides Multilingual Reasoning in Large Language Modelscited by 2
  6. 2026LLMs for Game Theory: Entropy-Guided In-Context Learning and Adaptive CoT Reasoningcited by 1
  7. 2026Agentic Conversational Search with Contextualized Reasoning via Reinforcement Learningcited by 1
  8. 2026Multi-Granularity Reasoning for Image Quality Assessment via Attribute-Aware Reinforcement Learning to Rankcited by 1
  9. 2026Efficient Test-Time Inference via Deterministic Exploration of Truncated Decoding Treescited by 1
  10. 2026CroSearch-R1: Better Leveraging Cross-lingual Knowledge for Retrieval-Augmented Generationcited by 1
  11. 2026Learning to Explore: Scaling Agentic Reasoning via Exploration-Aware Policy Optimizationcited by 1
  12. 2025WXImpactBench: A Disruptive Weather Impact Understanding Benchmark for Evaluating Large Language Modelscited by 0
  13. 2026LogitsCoder: Towards Efficient Chain-of-Thought Path Search via Logits Preference Decoding for Code Generationcited by 0
  14. 2026Avoiding Overthinking and Underthinking: Curriculum-Aware Budget Scheduling for LLMscited by 0
  15. 2026Entropy-informed Decoding: Adaptive Information-Driven Branchingcited by 0
  16. 2026BrailleBench: Investigating Multi-Criteria Braille Comprehension in Large Language Modelscited by 0
  17. 2026REWIND: Uncertainty-Guided Trajectory Repair for Test-Time Reasoningcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025Reasoning with Exploration: An Entropy Perspective
  2. 2025Explore Briefly, Then Decide: Mitigating LLM Overthinking via Cumulative Entropy Regulation
  3. 2025First Return, Entropy-Eliciting Explore
  4. 2026Entropy-Preserving Supervised Fine-Tuning via Adaptive Self-Distillation for Large Reasoning Models
  5. 2026ETR: Entropy Trend Reward for Efficient Chain-of-Thought Reasoning
  6. 2025PEAR: Phase Entropy Aware Reward for Efficient Reasoning
  7. 2026Taming the Thinker: Conditional Entropy Shaping for Adaptive LLM Reasoning
  8. 2025Making Slow Thinking Faster: Compressing LLM Chain-of-Thought via Step Entropy