mlsanityData to 13 Oct 2026

SCREWS: A Modular Framework for Reasoning with Revisions

arXiv 2309.13075 ↗First posted 20 Sep 2023cs.AI cs.CL cs.LGcited by 16

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

Abstract (arXiv)

Large language models (LLMs) can improve their accuracy on various tasks through iteratively refining and revising their output based on feedback. We observe that these revisions can introduce errors, in which case it is better to roll back to a previous result. Further, revisions are typically homogeneous: they use the same reasoning method that produced the initial answer, which may not correct errors. To enable exploration in this space, we present SCREWS, a modular framework for reasoning with revisions. It is comprised of three main modules: Sampling, Conditional Resampling, and Selection, each consisting of sub-modules that can be hand-selected per task. We show that SCREWS not only unifies several previous approaches under a common framework, but also reveals several novel strategies for identifying improved reasoning chains. We evaluate our framework with state-of-the-art LLMs (ChatGPT and GPT-4) on a diverse set of reasoning tasks and uncover useful new reasoning strategies for each: arithmetic word problems, multi-hop question answering, and code debugging. Heterogeneous revision strategies prove to be important, as does selection between original and revised candidates.

Key facts. Cited by 16 papers in the map (data through 13 Oct 2026). Placed in Test-time compute and efficient thinking, one of 17 directions in LLM reasoning. Builds on 44 of its 44 references that are in the map. First posted on arXiv 20 Sep 2023.

Builds on

44 of 44 references.

Closest ancestors · 4

Same direction, newest first

  1. 2023Tree of Thoughts: Deliberate Problem Solving with Large Language Models17 May
  2. 2023Self-Refine: Iterative Refinement with Self-Feedback30 Mar
  3. 2022Large Language Models Can Self-Improve20 Oct
  4. 2022Language Models (Mostly) Know What They Know11 Jul

Other references · 40

  1. 2017Attention Is All You Needcited by 60,106
  2. 2020Language Models are Few-Shot Learnerscited by 32,320
  3. 2023GPT-4 Technical Reportcited by 18,278
  4. 2022Chain-of-Thought Prompting Elicits Reasoning in Large Language Modelscited by 12,376
  5. 2021Training Verifiers to Solve Math Word Problemscited by 9,916
  6. 2022ReAct: Synergizing Reasoning and Acting in Language Modelscited by 8,502
  7. 2022Self-Consistency Improves Chain of Thought Reasoning in Language Modelscited by 5,657
  8. 2022Large Language Models are Zero-Shot Reasonerscited by 4,348
  9. 2023Toolformer: Language Models Can Teach Themselves to Use Toolscited by 4,117
  10. 2017Deep reinforcement learning from human preferencescited by 3,871
  11. 2022Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedbackcited by 3,691
  12. 2023Let's Verify Step by Stepcited by 3,672
28 more
  1. 2019Fine-Tuning Language Models from Human Preferencescited by 2,139
  2. 2022Beyond the Imitation Game: Quantifying and extrapolating the capabilities of language modelscited by 1,985
  3. 2022Least-to-Most Prompting Enables Complex Reasoning in Large Language Modelscited by 1,404
  4. 2022Solving Quantitative Reasoning Problems with Language Modelscited by 1,369
  5. 2022Program of Thoughts Prompting: Disentangling Computation from Reasoning for Numerical Reasoning Taskscited by 1,066
  6. 2023Teaching Large Language Models to Self-Debugcited by 972
  7. 2023Can LLMs Express Their Uncertainty? An Empirical Evaluation of Confidence Elicitation in LLMscited by 960
  8. 2023FrugalGPT: How to Use Large Language Models While Reducing Cost and Improving Performancecited by 815
  9. 2021Did Aristotle Use a Laptop? A Question Answering Benchmark with Implicit Reasoning Strategiescited by 738
  10. 2022Automatic Chain of Thought Prompting in Large Language Modelscited by 687
  11. 2023GPTScore: Evaluate as You Desirecited by 361
  12. 2023Check Your Facts and Try Again: Improving Large Language Models with External Knowledge and Automated Feedbackcited by 353
  13. 2022Large Language Models are Better Reasoners with Self-Verificationcited by 305
  14. 2022Teaching Small Language Models to Reasoncited by 282
  15. 2022Generating Sequences by Learning to Self-Correctcited by 273
  16. 2022Quark: Controllable Text Generation with Reinforced Unlearningcited by 237
  17. 2022Re3: Generating Longer Stories With Recursive Reprompting and Revisioncited by 193
  18. 2023LM vs LM: Detecting Factual Errors via Cross Examinationcited by 156
  19. 2022Distilling Reasoning Capabilities into Smaller Language Modelscited by 155
  20. 2023Progressive-Hint Prompting Improves Reasoning in Large Language Modelscited by 147
  21. 2023Deductive Verification of Chain-of-Thought Reasoningcited by 123
  22. 2019Multi-hop Reading Comprehension through Question Decomposition and Rescoringcited by 119
  23. 2023Question Decomposition Improves the Faithfulness of Model-Generated Reasoningcited by 108
  24. 2022Automatic Generation of Socratic Subquestions for Teaching Math Word Problemscited by 42
  25. 2023GRACE: Discriminator-Guided Chain-of-Thought Reasoningcited by 35
  26. 2021NL-EDIT: Correcting semantic parse errors through natural language interactioncited by 33
  27. 2023Natural Language Decomposition and Interpretation of Complex Utterancescited by 10
  28. 2021Interscript: A dataset for interactive learning of scripts through error feedbackcited by 9

Built on by

Cited by year (2026 to 13 Oct)

Citing papers per year: 2023: 5; 2024: 6; 2025: 4; 2026: 1.

16 papers.

Sort:

By area:

  1. 2023Navigate through Enigmatic Labyrinth A Survey of Chain of Thought Reasoning: Advances, Frontiers and Futurecited by 184
  2. 2023Pearl: Personalizing Large Language Model Writing Assistants with Generation-Calibrated Retrieverscited by 92
  3. 2024Calibrating Large Language Models with Sample Consistencycited by 73
  4. 2024LLM Self-Correction with DeCRIM: Decompose, Critique, and Refine for Enhanced Following of Instructions with Multiple Constraintscited by 35
  5. 2024DeepEdit: Knowledge Editing as Decoding with Constraintscited by 33
  6. 2023The ART of LLM Refinement: Ask, Refine, and Trustcited by 29
  7. 2023Mathematical Language Models: A Surveycited by 24
  8. 2024TOOLVERIFIER: Generalization to New Tools via Self-Verificationcited by 23
  9. 2024Mirror-Consistency: Harnessing Inconsistency in Majority Votingcited by 16
  10. 2025A Survey on Mathematical Reasoning and Optimization with Large Language Modelscited by 12
  11. 2026Revision or Re-Solving? Decomposing Second-Pass Gains in Multi-LLM Pipelinescited by 7
  12. 2023First-Step Advantage: Importance of Starting Right in Multi-Step Math Reasoningcited by 5
4 more
  1. 2024SMART: Self-learning Meta-strategy Agent for Reasoning Taskscited by 3
  2. 2025MARINE: Theoretical Optimization and Design for Multi-Agent Recursive IN-context Enhancementcited by 3
  3. 2025Reasoning Systems as Structured Processes: Foundations, Failures, and Formal Criteriacited by 1
  4. 2025EMAFusion: A Self-Optimizing System for Seamless LLM Selection and Integrationcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026A Single Revision Step Improves Token-Efficient LLM Reasoning
  2. 2025TurnBench-MS: A Benchmark for Evaluating Multi-Turn, Multi-Step Reasoning in Large Language Models
  3. 2026DeltaLogic: Minimal Premise Edits Reveal Belief-Revision Failures in Logical Reasoning Models
  4. 2024textttLM^texttt2: A Simple Society of Language Models Solves Complex Reasoning
  5. 2025Stop Overthinking: A Survey on Efficient Reasoning for Large Language Models
  6. 2026REVES: REvision and VErification--Augmented Training for Test-Time Scaling
  7. 2026TRACES: Tagging Reasoning Steps for Adaptive Cost-Efficient Early-Stopping
  8. 2024THOUGHTSCULPT: Reasoning with Intermediate Revision and Search