mlsanityData to 14 Oct 2026

Self-Polish: Enhance Reasoning in Large Language Models via Problem Refinement

arXiv 2305.14497 ↗First posted 23 May 2023cs.CL cs.AIcited by 43

On the map: LLM reasoning → Prompting and chain-of-thought reasoning

Abstract (arXiv)

To enhance the multi-step reasoning capabilities of large language models, researchers have extensively explored prompting methods, notably the Chain-of-Thought (CoT) method which explicitly elicits human-like rationales. However, they have inadvertently overlooked the potential of enhancing model reasoning performance by formulating higher-quality problems. In this work, we start from the problem side and propose Self-Polish (SP), a novel method that facilitates the model's reasoning by guiding it to progressively refine the given problems to be more comprehensible and solvable. We also explore several automatic prompting varients and propose the Self-Polish prompt bank for the community. SP is orthogonal to all other prompting methods of answer/reasoning side like CoT, allowing for seamless integration with state-of-the-art techniques for further improvement. Thorough experiments show that the proposed method attains notable and consistent effectiveness on five reasoning benchmarks across different models. Furthermore, our method also showcases impressive performance on robustness evaluation. Codes and prompts are available at https://github.com/WooooDyy/Self-Polish.

Key facts. Cited by 43 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 40 of its 65 references that are in the map. First posted on arXiv 23 May 2023.

Builds on

40 of 65 references.

Closest ancestors · 7

Same direction, newest first

  1. 2023Progressive-Hint Prompting Improves Reasoning in Large Language Models19 Apr
  2. 2022Reasoning with Language Model Prompting: A Survey19 Dec
  3. 2022Complexity-Based Prompting for Multi-Step Reasoning3 Oct
  4. 2022Large Language Models are Zero-Shot Reasoners24 May
  5. 2022Least-to-Most Prompting Enables Complex Reasoning in Large Language Models21 May
  6. 2022Self-Consistency Improves Chain of Thought Reasoning in Language Models21 Mar
  7. 2022Chain-of-Thought Prompting Elicits Reasoning in Large Language Models28 Jan

Other references · 33

  1. 2020Language Models are Few-Shot Learnerscited by 32,419
  2. 2022Training language models to follow instructions with human feedbackcited by 15,971
  3. 2021Training Verifiers to Solve Math Word Problemscited by 10,109
  4. 2021Measuring Mathematical Problem Solving With the MATH Datasetcited by 5,371
  5. 2022PaLM: Scaling Language Modeling with Pathwayscited by 5,326
  6. 2021On the Opportunities and Risks of Foundation Modelscited by 4,033
  7. 2021Finetuned Language Models Are Zero-Shot Learnerscited by 3,780
  8. 2022Scaling Instruction-Finetuned Language Modelscited by 3,024
  9. 2022Beyond the Imitation Game: Quantifying and extrapolating the capabilities of language modelscited by 1,998
  10. 2021Multitask Prompted Training Enables Zero-Shot Task Generalizationcited by 1,470
  11. 2022Rethinking the Role of Demonstrations: What Makes In-Context Learning Work?cited by 1,434
  12. 2022Solving Quantitative Reasoning Problems with Language Modelscited by 1,392
21 more
  1. 2018Continual Lifelong Learning with Neural Networks: A Reviewcited by 1,330
  2. 2021Scaling Language Models: Methods, Analysis & Insights from Training Gophercited by 1,202
  3. 2021What Makes Good In-Context Examples for GPT-3?cited by 1,195
  4. 2021Fantastically Ordered Prompts and Where to Find Them: Overcoming Few-Shot Prompt Order Sensitivitycited by 1,135
  5. 2022STaR: Bootstrapping Reasoning With Reasoningcited by 940
  6. 2021Are NLP Models really able to Solve Simple Math Word Problems?cited by 878
  7. 2023Large Language Models Can Be Easily Distracted by Irrelevant Contextcited by 841
  8. 2013An Empirical Investigation of Catastrophic Forgetting in Gradient-Based Neural Networkscited by 722
  9. 2019MathQA: Towards Interpretable Math Word Problem Solving with Operation-Based Formalismscited by 662
  10. 2017Program Induction by Rationale Generation : Learning to Solve and Explain Algebraic Word Problemscited by 650
  11. 2023Are Emergent Abilities of Large Language Models a Mirage?cited by 523
  12. 2016Solving General Arithmetic Word Problemscited by 360
  13. 2022Selection-Inference: Exploiting Large Language Models for Interpretable Logical Reasoningcited by 334
  14. 2018SWAG: A Large-Scale Adversarial Dataset for Grounded Commonsense Inferencecited by 333
  15. 2022Teaching Algorithmic Reasoning via In-context Learningcited by 112
  16. 2020Question Rewriting for Conversational Question Answeringcited by 93
  17. 2020Few-Shot Generative Conversational Query Rewritingcited by 85
  18. 2021CONQRR: Conversational Query Rewriting for Retrieval with Reinforcement Learningcited by 68
  19. 2020Conversational Question Reformulation via Sequence-to-Sequence Architectures and Pretrained Language Modelscited by 45
  20. 2018Identifying Well-formed Natural Language Questionscited by 29
  21. 2019How to Ask Better Questions? A Large-Scale Multi-Domain Dataset for Rewriting Ill-Formed Questionscited by 12

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2023: 4; 2024: 24; 2025: 13; 2026: 2.

43 papers.

Sort:

By area:

  1. 2023The Rise and Potential of Large Language Model Based Agents: A Surveycited by 1,439
  2. 2024HuatuoGPT-o1, Towards Medical Complex Reasoning with LLMscited by 265
  3. 2024Secrets of RLHF in Large Language Models Part II: Reward Modelingcited by 171
  4. 2024Rethinking the Bounds of LLM Reasoning: Are Multi-Agent Discussions the Key?cited by 159
  5. 2025LLM Post-Training: A Deep Dive into Reasoning Large Language Modelscited by 135
  6. 2023Large Language Models for Robotics: A Surveycited by 133
  7. 2024A Review of Large Language Models and Autonomous Agents in Chemistrycited by 99
  8. 2024Self-Contrast: Better Reflection Through Inconsistent Solving Perspectivescited by 71
  9. 2024Training Large Language Models for Reasoning through Reverse Curriculum Reinforcement Learningcited by 65
  10. 2024Can Language Models Perform Robust Reasoning in Chain-of-thought Prompting with Noisy Rationales?cited by 41
  11. 2024E-SQL: Direct Schema Linking via Question Enrichment in Text-to-SQLcited by 35
  12. 2024Confidence v.s. Critique: A Decomposition of Self-Correction Capability for LLMscited by 29
31 more
  1. 2024Instance-adaptive Zero-shot Chain-of-Thought Promptingcited by 26
  2. 2023Lyra: Orchestrating Dual Correction in Automated Theorem Provingcited by 25
  3. 2025From Passive to Active Reasoning: Can Large Language Models Ask the Right Questions under Incomplete Information?cited by 23
  4. 2024MetaReflection: Learning Instructions for Language Agents using Past Reflectionscited by 19
  5. 2024Counterfactual Debating with Preset Stances for Hallucination Elimination of LLMscited by 17
  6. 2024Mirror-Consistency: Harnessing Inconsistency in Majority Votingcited by 17
  7. 2024Exploring the Compositional Deficiency of Large Language Models in Mathematical Reasoningcited by 15
  8. 2025CompileAgent: Automated Real-World Repo-Level Compilation with Tool-Integrated LLM-based Agent Systemcited by 15
  9. 2024Understanding the Effects of Iterative Prompting on Truthfulnesscited by 14
  10. 2025Instruct-of-Reflection: Enhancing Large Language Models Iterative Reflection Capabilities via Dynamic-Meta Instructioncited by 10
  11. 2025Self-Harmony: Learning to Harmonize Self-Supervision and Self-Play in Test-Time Reinforcement Learningcited by 10
  12. 2024Affordable Generative Agentscited by 9
  13. 2024Enhancing the General Agent Capabilities of Low-Parameter LLMs through Tuning and Multi-Branch Reasoningcited by 9
  14. 2025A Probabilistic Inference Scaling Theory for LLM Self-Correctioncited by 9
  15. 2024DataLab: A Unified Platform for LLM-Powered Business Intelligencecited by 8
  16. 2025GenoMAS: A Multi-Agent Framework for Scientific Discovery via Code-Driven Gene Expression Analysiscited by 8
  17. 2023On the Discussion of Large Language Models: Symmetry of Agents and Interplay with Promptscited by 5
  18. 2024GenoTEX: An LLM Agent Benchmark for Automated Gene Expression Data Analysiscited by 5
  19. 2024Dr3: Ask Large Language Models Not to Give Off-Topic Answers in Open Domain Multi-Hop Question Answeringcited by 4
  20. 2025SEED: Enhancing Text-to-SQL Performance and Practical Usability Through Automatic Evidence Generationcited by 4
  21. 2026FACT-E: Causality-Inspired Evaluation for Trustworthy Chain-of-Thought Reasoningcited by 4
  22. 2024Revolutionizing Bridge Operation and Maintenance with LLM-based Agents: An Overview of Applications and Insightscited by 3
  23. 2025From Belief Entrenchment to Robust Reasoning in LLM Agentscited by 3
  24. 2024Can Separators Improve Chain-of-Thought Prompting?cited by 2
  25. 2024360^circREA: Towards A Reusable Experience Accumulation with 360deg Assessment for Multi-Agent Systemcited by 2
  26. 2025Multi-Stakeholder Alignment in LLM-Powered Collaborative AI Systems: A Multi-Agent Framework for Intelligent Tutoringcited by 2
  27. 2025Knowledge-Augmented Long-CoT Generation for Complex Biomolecular Reasoningcited by 2
  28. 2025On the Convergence of Moral Self-Correction in Large Language Modelscited by 1
  29. 2026iFlip: Iterative Feedback-driven Counterfactual Example Refinementcited by 1
  30. 2024Enhancing Knowledge Distillation for LLMs with Response-Priming Promptingcited by 0
  31. 2025OmniReflect: Discovering Transferable Constitutions for LLM agents via Neuro-Symbolic Reflectionscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2023Just Ask One More Time! Self-Agreement Improves Reasoning of Language Models in (Almost) All Scenarios
  2. 2024Auto-Evolve: Enhancing Large Language Model's Performance via Self-Reasoning Framework
  3. 2025PromptCoT 2.0: Scaling Prompt Synthesis for Large Language Model Reasoning
  4. 2025Self-Supervised Prompt Optimization
  5. 2022Large Language Models are Better Reasoners with Self-Verification
  6. 2025Advancing Multi-Step Mathematical Reasoning in Large Language Models through Multi-Layered Self-Reflection with Auto-Prompting
  7. 2022Large Language Models Can Self-Improve
  8. 2024Self-Discover: Large Language Models Self-Compose Reasoning Structures