mlsanityData to 14 Oct 2026

CodeI/O: Condensing Reasoning Patterns via Code Input-Output Prediction

arXiv 2502.07316 ↗First posted 11 Feb 2025cs.CL cs.AIcited by 55

On the map: LLM reasoning → Reasoning for agents, planning and code

Abstract (arXiv)

Reasoning is a fundamental capability of Large Language Models. While prior research predominantly focuses on enhancing narrow skills like math or code generation, improving performance on many other reasoning tasks remains challenging due to sparse and fragmented training data. To address this issue, we propose CodeI/O, a novel approach that systematically condenses diverse reasoning patterns inherently embedded in contextually-grounded codes, through transforming the original code into a code input-output prediction format. By training models to predict inputs/outputs given code and test cases entirely in natural language as Chain-of-Thought (CoT) rationales, we expose them to universal reasoning primitives -- like logic flow planning, state-space searching, decision tree traversal, and modular decomposition -- while decoupling structured reasoning from code-specific syntax and preserving procedural rigor. Experimental results demonstrate CodeI/O leads to consistent improvements across symbolic, scientific, logic, math & numerical, and commonsense reasoning tasks. By matching the existing ground-truth outputs or re-executing the code with predicted inputs, we can verify each prediction and further enhance the CoTs through multi-turn revision, resulting in CodeI/O++ and achieving higher performance. Our data and models are available at https://github.com/hkust-nlp/CodeIO.

Key facts. Cited by 55 papers in the map (data through 14 Oct 2026). Placed in Reasoning for agents, planning and code, one of 17 directions in LLM reasoning. Builds on 34 of its 52 references that are in the map. First posted on arXiv 11 Feb 2025.

Builds on

34 of 52 references.

Closest ancestors · 3

Same direction, newest first

  1. 2024OpenAI o1 System Card21 Dec
  2. 2024Qwen2.5-Coder Technical Report18 Sep
  3. 2024LiveCodeBench: Holistic and Contamination Free Evaluation of Large Language Models for Code12 Mar

Other references · 31

  1. 2024The Llama 3 Herd of Modelscited by 14,715
  2. 2021Training Verifiers to Solve Math Word Problemscited by 10,109
  3. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 9,379
  4. 2025DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learningcited by 6,330
  5. 2023Qwen Technical Reportcited by 3,344
  6. 2023Code Llama: Open Foundation Models for Codecited by 2,139
  7. 2022Beyond the Imitation Game: Quantifying and extrapolating the capabilities of language modelscited by 1,998
  8. 2024Gemma 2: Improving Open Language Models at a Practical Sizecited by 1,998
  9. 2024Mixtral of Expertscited by 1,823
  10. 2024DeepSeek-V2: A Strong, Economical, and Efficient Mixture-of-Experts Language Modelcited by 1,135
  11. 2024Qwen2.5-Math Technical Report: Toward Mathematical Expert Model via Self-Improvementcited by 1,090
  12. 2024Tulu 3: Pushing Frontiers in Open Language Model Post-Trainingcited by 1,053
19 more
  1. 2021Show Your Work: Scratchpads for Intermediate Computation with Language Modelscited by 886
  2. 2014Neural Turing Machinescited by 644
  3. 2022Towards Reasoning in Large Language Models: A Surveycited by 596
  4. 2024StarCoder 2 and The Stack v2: The Next Generationcited by 502
  5. 2023Scaling Relationship on Learning Mathematical Reasoning with Large Language Modelscited by 427
  6. 2024DeepSeek-Coder-V2: Breaking the Barrier of Closed-Source Models in Code Intelligencecited by 331
  7. 2024LiveBench: A Challenging, Contamination-Limited LLM Benchmarkcited by 262
  8. 2022Reasoning with Language Model Prompting: A Surveycited by 252
  9. 2025ZebraLogic: On the Scaling Limits of LLMs for Logical Reasoningcited by 201
  10. 2024OpenMathInstruct-2: Accelerating AI for Math with Massive Open-Source Instruction Datacited by 200
  11. 2024MAmmoTH2: Scaling Instructions from the Webcited by 145
  12. 2024OpenCoder: The Open Cookbook for Top-Tier Code Large Language Modelscited by 130
  13. 2024InternLM-Math: Open Math Large Language Models Toward Verifiable Reasoningcited by 119
  14. 2025Towards System 2 Reasoning in LLMs: Learning How to Think With Meta Chain-of-Thoughtcited by 109
  15. 2014Learning to Executecited by 106
  16. 2024KOR-Bench: Benchmarking Language Models on Knowledge-Orthogonal Reasoning Taskscited by 70
  17. 2024Evaluation data contamination in LLMs: how do we measure it and (when) does it matter?cited by 32
  18. 2024Skywork-Math: Data Scaling Laws for Mathematical Reasoning in Large Language Models -- The Story Goes Oncited by 24
  19. 2024LogicPro: Improving Complex Logical Reasoning via Program-Guided Learningcited by 14

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 28; 2026: 27.

55 papers.

Sort:

By area:

  1. 2025Towards Reasoning Era: A Survey of Long Chain-of-Thought for Reasoning Large Language Modelscited by 384
  2. 2025Absolute Zero: Reinforced Self-play Reasoning with Zero Datacited by 344
  3. 2025Dream-Coder 7B: An Open Diffusion Language Model for Codecited by 81
  4. 2025Revisiting Reinforcement Learning for LLM Reasoning from A Cross-Domain Perspectivecited by 76
  5. 20251.4 Million Open-Source Distilled Reasoning Dataset to Empower Large Language Model Trainingcited by 49
  6. 2025ARM: Adaptive Reasoning Modelcited by 39
  7. 2025CodeRL+: Improving Code Generation via Reinforcement with Execution Semantics Alignmentcited by 24
  8. 2025O2-Searcher: A Searching-based Agent Model for Open-Domain Open-Ended Question Answeringcited by 23
  9. 2025Could Thinking Multilingually Empower LLM Reasoning?cited by 16
  10. 2026NeuReasoner: Towards Explainable, Controllable, and Unified Reasoning via Mixture-of-Neuronscited by 14
  11. 2025Breaking the Data Barrier -- Building GUI Agents Through Task Generalizationcited by 13
  12. 2025Scaling Code-Assisted Chain-of-Thoughts and Instructions for Model Reasoningcited by 10
38 more
  1. 2025Youtu-LLM: Unlocking the Native Agentic Potential for Lightweight Large Language Modelscited by 10
  2. 2026Can We Predict Before Executing Machine Learning Agents?cited by 10
  3. 2025Variational Reasoning for Language Modelscited by 8
  4. 2025Reasoning Curriculum: Bootstrapping Broad LLM Reasoning from Mathcited by 8
  5. 2025Evaluating the Generalization Capabilities of Large Language Models on Code Reasoningcited by 7
  6. 2025Which Data Attributes Stimulate Math and Code Reasoning? An Investigation via Influence Functionscited by 7
  7. 2025Think in Games: Learning to Reason in Games via Reinforcement Learning with Large Language Modelscited by 7
  8. 2025LLMs Lean on Priors, Not Programming Language Semanticscited by 6
  9. 2026ExecVerify: White-Box RL with Verifiable Stepwise Rewards for Code Execution Reasoningcited by 6
  10. 2026Learning, Fast and Slow: Towards LLMs That Adapt Continuallycited by 6
  11. 2026From Chatbot to Digital Colleague: The Paradigm Shift Toward Persistent Autonomous AIcited by 6
  12. 2026Ling and Ring 2.6 Technical Report: Efficient and Instant Agentic Intelligence at Trillion-Parameter Scalecited by 6
  13. 2025Rethinking Reasoning Quality in Large Language Models through Enhanced Chain-of-Thought via RLcited by 5
  14. 2025RE-Searcher: Robust Agentic Search with Goal-oriented Planning and Self-reflectioncited by 5
  15. 2026Your Language Model is Its Own Critic: Reinforcement Learning with Value Estimation from Actor's Internal Statescited by 5
  16. 2026Transferability for General Reasoning: An Automated Curriculum for Multi-Domain RLVRcited by 5
  17. 2026Verify Before You Distill: Prompt-Level Teacher Gating for On-Policy Distillationcited by 5
  18. 2025The Hitchhiker's Guide to Program Analysis, Part II: Deep Thoughts by LLMscited by 4
  19. 2025CodeReasoner: Enhancing the Code Reasoning Ability with Reinforcement Learningcited by 4
  20. 2025QueST: Incentivizing LLMs to Generate Difficult Problemscited by 4
  21. 2026Can LLMs Compress (and Decompress)? Evaluating Code Understanding and Execution via Invertibilitycited by 4
  22. 2026ZAYA1-8B Technical Reportcited by 4
  23. 2026Role-Agent: Bootstrapping LLM Agents via Dual-Role Evolutioncited by 4
  24. 2025SURGE: On the Potential of Large Language Models as General-Purpose Surrogate Code Executorscited by 3
  25. 2025Teaching LLM to Reason: Reinforcement Learning from Algorithmic Problems without Codecited by 3
  26. 2025What-If Analysis of Large Language Models: Explore the Game World Using Proactive Thinkingcited by 3
  27. 2026Self-Execution Simulation Improves Coding Modelscited by 3
  28. 2026StepCodeReasoner: Aligning Code Reasoning with Stepwise Execution Traces via Reinforcement Learningcited by 2
  29. 2026D3-MOPD: Dynamic Domain ScheDuling for Efficient Multi-Teacher Distillationcited by 2
  30. 2025R1-Fuzz: Specializing Language Models for Textual Fuzzing via Reinforcement Learningcited by 1
  31. 2026Can Large Language Models Generalize Procedures Across Representations?cited by 1
  32. 2026CodePivot: Bootstrapping Multilingual Transpilation in LLMs via Reinforcement Learning without Parallel Corporacited by 1
  33. 2026Enhancing the Code Reasoning Capabilities of LLMs via Consistency-based Reinforcement Learningcited by 1
  34. 2026Le Critique: Privileged Value Functions for LLM Reinforcement Learningcited by 1
  35. 2025Generating Verifiable Chain of Thoughts from Exection-Tracescited by 0
  36. 2026How Reasoning Evolves from Post-Training Data: An Empirical Study Using Chesscited by 0
  37. 2026Agent-GWO: Collaborative Agents for Dynamic Prompt Optimization in Large Language Modelscited by 0
  38. 2026Pseudocode-Guided Structured Reasoning for Automating Reliable Inference in Vision-Language Modelscited by 0

50 most cited of 55 All 55 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2026Exploring Extrinsic and Intrinsic Properties for Effective Reasoning with Code Interpreter
  2. 2026Reasoning Core: A Scalable Procedural Data Generation Suite for Symbolic Pre-training and Post-Training
  3. 2025CoT-RAG: Integrating Chain of Thought and Retrieval-Augmented Generation to Enhance Reasoning in Large Language Models
  4. 2025CoRT: Code-integrated Reasoning within Thinking
  5. 2025Code Execution as Grounded Supervision for LLM Reasoning
  6. 2025CodeSense: a Real-World Benchmark and Dataset for Code Semantic Reasoning
  7. 2025Unveiling the Magic of Code Reasoning through Hypothesis Decomposition and Amendment
  8. 2023Code Prompting: a Neural Symbolic Method for Complex Reasoning in Large Language Models