mlsanityData to 7 Oct 2026

OpenMathInstruct-2: Accelerating AI for Math with Massive Open-Source Instruction Data

arXiv 2410.01560 ↗First posted 2 Oct 2024cs.CL cs.AI cs.LGcited by 194

On the map: LLM reasoning → Mathematical reasoning

Abstract (arXiv)

Mathematical reasoning continues to be a critical challenge in large language model (LLM) development with significant interest. However, most of the cutting-edge progress in mathematical reasoning with LLMs has become closed-source due to lack of access to training data. This lack of data access limits researchers from understanding the impact of different choices for synthesizing and utilizing the data. With the goal of creating a high-quality finetuning (SFT) dataset for math reasoning, we conduct careful ablation experiments on data synthesis using the recently released textttLlama3.1 family of models. Our experiments show that: (a) solution format matters, with excessively verbose solutions proving detrimental to SFT performance, (b) data generated by a strong teacher outperforms equally-sized data generated by a weak student model, (c) SFT is robust to low-quality solutions, allowing for imprecise data filtering, and (d) question diversity is crucial for achieving data scaling gains. Based on these insights, we create the OpenMathInstruct-2 dataset, which consists of 14M question-solution pairs (≈ 600K unique questions), making it nearly eight times larger than the previous largest open-source math reasoning dataset. Finetuning the textttLlama-3.1-8B-Base using OpenMathInstruct-2 outperforms textttLlama3.1-8B-Instruct on MATH by an absolute 15.9% (51.9% rightarrow 67.8%). Finally, to accelerate the open-source efforts, we release the code, the finetuned models, and the OpenMathInstruct-2 dataset under a commercially permissive license.

Key facts. Cited by 194 papers in the map (data through 7 Oct 2026). Placed in Mathematical reasoning, one of 17 directions in LLM reasoning. Builds on 25 of its 39 references that are in the map. First posted on arXiv 2 Oct 2024.

Builds on

25 of 39 references.

Closest ancestors · 8

Same direction, newest first

  1. 2024InfinityMATH: A Scalable Instruction Tuning Dataset in Programmatic Mathematical Reasoning9 Aug
  2. 2024Skywork-Math: Data Scaling Laws for Mathematical Reasoning in Large Language Models -- The Story Goes On11 Jul
  3. 2024A Careful Examination of Large Language Model Performance on Grade School Arithmetic1 May
  4. 2024Common 7B Language Models Already Possess Strong Math Capabilities7 Mar
  5. 2024OpenMathInstruct-1: A 1.8 Million Math Instruction Tuning Dataset15 Feb
  6. 2023Llemma: An Open Language Model For Mathematics16 Oct
  7. 2023MuggleMath: Assessing the Impact of Query and Response Augmentation on Math Reasoning9 Oct
  8. 2023ToRA: A Tool-Integrated Reasoning Agent for Mathematical Problem Solving29 Sep

Other references · 17

  1. 2022Chain-of-Thought Prompting Elicits Reasoning in Large Language Modelscited by 12,302
  2. 2021Training Verifiers to Solve Math Word Problemscited by 9,743
  3. 2022Self-Consistency Improves Chain of Thought Reasoning in Language Modelscited by 5,601
  4. 2019The Curious Case of Neural Text Degenerationcited by 2,881
  5. 2022Rethinking the Role of Demonstrations: What Makes In-Context Learning Work?cited by 1,413
  6. 2022Program of Thoughts Prompting: Disentangling Computation from Reasoning for Numerical Reasoning Taskscited by 1,057
  7. 2024Large Language Monkeys: Scaling Inference Compute with Repeated Samplingcited by 946
  8. 2021Show Your Work: Scratchpads for Intermediate Computation with Language Modelscited by 871
  9. 2023MetaMath: Bootstrap Your Own Mathematical Questions for Large Language Modelscited by 724
  10. 2023MAmmoTH: Building Math Generalist Models through Hybrid Instruction Tuningcited by 483
  11. 2024Dolma: an Open Corpus of Three Trillion Tokens for Language Model Pretraining Researchcited by 475
  12. 2024DeepSeek-Coder-V2: Breaking the Barrier of Closed-Source Models in Code Intelligencecited by 325
5 more
  1. 2023Rethinking Benchmark and Contamination for Language Models with Rephrased Samplescited by 192
  2. 2024MAmmoTH2: Scaling Instructions from the Webcited by 143
  3. 2023Can we trust the evaluation on ChatGPT?cited by 63
  4. 2024Smaller, Weaker, Yet Better: Training LLM Reasoners via Compute-Optimal Samplingcited by 60
  5. 2023LLM-Assisted Code Cleaning For Training Accurate Code Generatorscited by 27

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 10; 2025: 109; 2026: 75.

194 papers.

Sort:

By area:

  1. 2024Tulu 3: Pushing Frontiers in Open Language Model Post-Trainingcited by 988
  2. 2024DeepSeek-VL2: Mixture-of-Experts Vision-Language Models for Advanced Multimodal Understandingcited by 559
  3. 2025Process Reinforcement through Implicit Rewardscited by 478
  4. 2025Towards Reasoning Era: A Survey of Long Chain-of-Thought for Reasoning Large Language Modelscited by 381
  5. 2025Scaling up Test-Time Compute with Latent Reasoning: A Recurrent Depth Approachcited by 372
  6. 2025rStar-Math: Small LLMs Can Master Math Reasoning with Self-Evolved Deep Thinkingcited by 308
  7. 2025OpenThoughts: Data Recipes for Reasoning Modelscited by 287
  8. 2025DeepMath-103K: A Large-Scale, Challenging, Decontaminated, and Verifiable Mathematical Dataset for Advancing Reasoningcited by 261
  9. 2025Light-R1: Curriculum SFT, DPO and RL for Long COT from Scratch and Beyondcited by 175
  10. 2025AIMO-2 Winning Solution: Building State-of-the-Art Mathematical Reasoning Models with OpenMathReasoning datasetcited by 165
  11. 2025Llama-Nemotron: Efficient Reasoning Modelscited by 151
  12. 2025Echo Chamber: RL Post-training Amplifies Behaviors Learned in Pretrainingcited by 123
38 more
  1. 2025KodCode: A Diverse, Challenging, and Verifiable Synthetic Dataset for Codingcited by 115
  2. 2025Small Models Struggle to Learn from Strong Reasonerscited by 109
  3. 2025Towards System 2 Reasoning in LLMs: Learning How to Think With Meta Chain-of-Thoughtcited by 108
  4. 2025NVIDIA Nemotron Nano 2: An Accurate and Efficient Hybrid Mamba-Transformer Reasoning Modelcited by 94
  5. 2025Nemotron-H: A Family of Accurate and Efficient Hybrid Mamba-Transformer Modelscited by 88
  6. 2025Nemotron 3 Nano: Open, Efficient Mixture-of-Experts Hybrid Mamba-Transformer Model for Agentic Reasoningcited by 85
  7. 2025NaturalReasoning: Reasoning in the Wild with 2.8M Challenging Questionscited by 81
  8. 2024AceMath: Advancing Frontier Math Reasoning with Post-Training and Reward Modelingcited by 76
  9. 2025Self-rewarding correction for mathematical reasoningcited by 67
  10. 2025Dion: Distributed Orthonormalized Updatescited by 64
  11. 2025Critique Fine-Tuning: Learning to Critique is More Effective than Learning to Imitatecited by 61
  12. 2025MegaScience: Pushing the Frontiers of Post-Training Datasets for Science Reasoningcited by 56
  13. 2024A Survey of Mathematical Reasoning in the Era of Multimodal Large Language Model: Benchmark, Method & Challengescited by 55
  14. 2025CodeI/O: Condensing Reasoning Patterns via Code Input-Output Predictioncited by 54
  15. 2025Satori: Reinforcement Learning with Chain-of-Action-Thought Enhances LLM Reasoning via Autoregressive Searchcited by 53
  16. 2025A Survey on Large Language Models for Mathematical Reasoningcited by 52
  17. 2025FineVision: Open Data Is All You Needcited by 51
  18. 2025MegaMath: Pushing the Limits of Open Math Corporacited by 47
  19. 2025TaskCraft: Automated Generation of Agentic Taskscited by 46
  20. 2025UGPhysics: A Comprehensive Benchmark for Undergraduate Physics Reasoning with Large Language Modelscited by 45
  21. 2025Scaling Laws of Synthetic Data for Language Modelscited by 44
  22. 2025Understand, Solve and Translate: Bridging the Multilingual Mathematical Reasoning Gapcited by 41
  23. 2025Universal Cross-Tokenizer Distillation via Approximate Likelihood Matchingcited by 40
  24. 2025RL Tango: Reinforcing Generator and Verifier Together for Language Reasoningcited by 39
  25. 2026HDPO: Hybrid Distillation Policy Optimization via Privileged Self-Distillationcited by 39
  26. 2024Entropy-Regularized Process Reward Modelcited by 38
  27. 2025Trust, But Verify: A Self-Verification Approach to Reinforcement Learning with Verifiable Rewardscited by 38
  28. 2025Prismatic Synthesis: Gradient-based Data Diversification Boosts Generalization in LLM Reasoningcited by 38
  29. 2025S2R: Teaching LLMs to Self-verify and Self-correct via Reinforcement Learningcited by 37
  30. 2025Large-Scale Data Selection for Instruction Tuningcited by 34
  31. 2025Optimizing Temperature for Language Models with Multi-Sample Inferencecited by 33
  32. 2025Mitigating Overthinking in Large Reasoning Models via Manifold Steeringcited by 32
  33. 2025The First Few Tokens Are All You Need: An Efficient and Effective Unsupervised Prefix Fine-Tuning Method for Reasoning Modelscited by 31
  34. 2025Mavors: Multi-granularity Video Representation for Multimodal Large Language Modelcited by 31
  35. 2025MathFusion: Enhancing Mathematical Problem-solving of LLM through Instruction Fusioncited by 30
  36. 2025Thinking Slow, Fast: Scaling Inference Compute with Distilled Reasonerscited by 22
  37. 2025Nemotron-Math: Efficient Long-Context Distillation of Mathematical Reasoning from Multi-Mode Supervisioncited by 22
  38. 2025M1: Towards Scalable Test-Time Compute with Mamba Reasoning Modelscited by 21

50 most cited of 194 All 194 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2025JT-Math: A Multi-Stage Framework for Advanced Mathematical Reasoning in Large Language Models
  2. 2024MathScale: Scaling Instruction Tuning for Mathematical Reasoning
  3. 2025Mathematical Computation and Reasoning Errors by Large Language Models
  4. 2025Advancing Mathematical Reasoning in Language Models: The Impact of Problem-Solving Data, Data Synthesis Methods, and Training Stages
  5. 2024MindStar: Enhancing Math Reasoning in Pre-trained LLMs at Inference Time
  6. 2024Benchmarking Large Language Models for Math Reasoning Tasks
  7. 2024JiuZhang3.0: Efficiently Improving Mathematical Reasoning by Training Small Data Synthesis Models
  8. 2026More Accurate or More Efficient? Evaluating Locally Deployed Compact Open-Weight Language Models for Mathematical Reasoning