mlsanityData to 7 Oct 2026

CoT-Self-Instruct: Building high-quality synthetic prompts for reasoning and non-reasoning tasks

arXiv 2507.23751 ↗First posted 31 Jul 2025cs.AI cs.CLcited by 30

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

also in 1 more
Abstract (arXiv)

We propose CoT-Self-Instruct, a synthetic data generation method that instructs LLMs to first reason and plan via Chain-of-Thought (CoT) based on given seed tasks, and then generate a new synthetic example of similar quality and complexity. This is followed by a filtering step to select high-quality data using automatic metrics, which are then used for LLM training. In verifiable reasoning, our synthetic data significantly outperforms existing training datasets, such as s1k and OpenMathReasoning, when evaluated on MATH500, AMC23, AIME24, and GPQA-Diamond. For non-verifiable instruction-following tasks, our method surpasses the performance of both human and standard Self-Instruct training data on the AlpacaEval 2.0 and Arena-Hard benchmarks.

Key facts. Cited by 30 papers in the map (data through 7 Oct 2026). Placed in Prompting and chain-of-thought reasoning, one of 17 directions in LLM reasoning. Builds on 55 of its 57 references that are in the map. First posted on arXiv 31 Jul 2025.

Builds on

55 of 57 references.

References · 55

  1. 2024The Llama 3 Herd of Modelscited by 14,192
  2. 2023Llama 2: Open Foundation and Fine-Tuned Chat Modelscited by 12,865
  3. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 8,851
  4. 2025Qwen3 Technical Reportcited by 8,195
  5. 2023Direct Preference Optimization: Your Language Model is Secretly a Reward Modelcited by 7,916
  6. 2025DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learningcited by 6,146
  7. 2023Efficient Memory Management for Large Language Model Serving with PagedAttentioncited by 5,767
  8. 2021Measuring Mathematical Problem Solving With the MATH Datasetcited by 5,203
  9. 2021Finetuned Language Models Are Zero-Shot Learnerscited by 3,747
  10. 2023Let's Verify Step by Stepcited by 3,598
  11. 2023GPQA: A Graduate-Level Google-Proof Q&A Benchmarkcited by 2,748
  12. 2022Self-Instruct: Aligning Language Models with Self-Generated Instructionscited by 2,674
38 more
  1. 2020Learning to summarize from human feedbackcited by 2,562
  2. 2024HybridFlow: A Flexible and Efficient RLHF Frameworkcited by 1,980
  3. 2025s1: Simple test-time scalingcited by 1,311
  4. 2024Length-Controlled AlpacaEval: A Simple Way to Debias Automatic Evaluatorscited by 865
  5. 2022Super-NaturalInstructions: Generalization via Declarative Instructions on 1600+ NLP Taskscited by 858
  6. 2023ChatGPT Outperforms Crowd-Workers for Text-Annotation Taskscited by 849
  7. 2023Enhancing Chat Language Models by Scaling High-quality Instructional Conversationscited by 825
  8. 2021Cross-Task Generalization via Natural Language Crowdsourcing Instructionscited by 696
  9. 2024Phi-4 Technical Reportcited by 675
  10. 2024Self-Rewarding Language Modelscited by 654
  11. 2024WildChat: 1M ChatGPT Interaction Logs in the Wildcited by 598
  12. 2025LIMO: Less is More for Reasoningcited by 488
  13. 2024From Crowdsourced Data to High-Quality Benchmarks: Arena-Hard and BenchBuilder Pipelinecited by 484
  14. 2023How Close is ChatGPT to Human Experts? Comparison Corpus, Evaluation, and Detectioncited by 472
  15. 2023Scaling Relationship on Learning Mathematical Reasoning with Large Language Modelscited by 417
  16. 2023What Makes Good Data for Alignment? A Comprehensive Study of Automatic Data Selection in Instruction Tuningcited by 354
  17. 2019Analysing Mathematical Reasoning Abilities of Neural Modelscited by 331
  18. 2023Beyond Human Data: Scaling Self-Training for Problem-Solving with Language Modelscited by 325
  19. 2025Absolute Zero: Reinforced Self-play Reasoning with Zero Datacited by 323
  20. 2022Language Models are Realistic Tabular Data Generatorscited by 290
  21. 2024On LLMs-Driven Synthetic Data Generation, Curation, and Evaluation: A Surveycited by 240
  22. 2024Meta-Rewarding Language Models: Self-Improving Alignment with LLM-as-a-Meta-Judgecited by 202
  23. 2022Generating Training Data with Language Models: Towards Zero-Shot Language Understandingcited by 199
  24. 2025AIMO-2 Winning Solution: Building State-of-the-Art Mathematical Reasoning Models with OpenMathReasoning datasetcited by 165
  25. 2024Rephrasing the Web: A Recipe for Compute and Data-Efficient Language Modelingcited by 156
  26. 2023Self-Alignment with Instruction Backtranslationcited by 149
  27. 2023Generative agent-based modeling with actions grounded in physical, social, or digital space using Concordiacited by 119
  28. 2023#InsTag: Instruction Tagging for Analyzing Supervised Fine-tuning of Large Language Modelscited by 108
  29. 2023A Complete Survey on Generative AI (AIGC): Is ChatGPT from GPT-4 to GPT-5 All You Need?cited by 84
  30. 2023Human Feedback is not Gold Standardcited by 82
  31. 2025NaturalReasoning: Reasoning in the Wild with 2.8M Challenging Questionscited by 81
  32. 2023InstructZero: Efficient Instruction Optimization for Black-Box Large Language Modelscited by 68
  33. 2023Harnessing large-language models to generate private synthetic textcited by 65
  34. 2025Self-Challenging Language Model Agentscited by 56
  35. 2024Self-Consistency Preference Optimizationcited by 50
  36. 2024Better Synthetic Data by Retrieving and Transforming Existing Datasetscited by 36
  37. 2025Recycling the Web: A Method to Enhance Pre-training Data Quality and Quantity for Language Modelscited by 31
  38. 2024Source2Synth: Synthetic Data Generation and Curation Grounded in Real Data Sourcescited by 27

50 most cited of 55 All 55 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 11; 2026: 19.

30 papers.

Sort:

By area:

  1. 2025SPICE: Self-Play In Corpus Environments Improves Reasoningcited by 78
  2. 2025DataFlow: An LLM-Driven Framework for Unified Data Preparation and Workflow Automation in the Era of Data-Centric AIcited by 22
  3. 2026From Self-Evolving Synthetic Data to Verifiable-Reward RL: Post-Training Multi-turn Interactive Tool-Using Agentscited by 17
  4. 2025RESTRAIN: From Spurious Votes to Signals -- Self-Driven RL with Self-Penalizationcited by 13
  5. 2026Autodata: An agentic data scientist to create high quality synthetic datacited by 12
  6. 2026Agentic Proposing: Enhancing Large Language Model Reasoning via Compositional Skill Synthesiscited by 11
  7. 2025Learning to Pose Problems: Reasoning-Driven and Solver-Adaptive Data Synthesiscited by 8
  8. 2025AI & Human Co-Improvement for Safer Co-Superintelligencecited by 8
  9. 2025Enhancing Large Language Model Reasoning with Reward Models: An Analytical Surveycited by 7
  10. 2025DESIGNER: Design-Logic-Guided Multidisciplinary Data Synthesis for LLM Reasoningcited by 6
  11. 2025ScaleDiff: Scaling Difficult Problems for Advanced Mathematical Reasoningcited by 6
  12. 2026Skill Self-Play: Pushing the Frontier of LLM Capability with Co-Evolving Skillscited by 6
18 more
  1. 2026Self-Improvement of Large Language Models: A Technical Overview and Future Outlookcited by 5
  2. 2025Tailored Primitive Initialization is the Secret Key to Reinforcement Learningcited by 4
  3. 2026DS2-Instruct: Domain-Specific Data Synthesis for Large Language Models Instruction Tuningcited by 4
  4. 2026On the Generalization Gap in Self-Evolving Language Model Reasoningcited by 4
  5. 2025Towards Label-Free Biological Reasoning Synthetic Dataset Creation via Uncertainty Filteringcited by 3
  6. 2025Beyond MCQ: An Open-Ended Arabic Cultural QA Benchmark with Dialect Variantscited by 3
  7. 2026Reasoning over mathematical objects: on-policy reward modeling and test time aggregationcited by 3
  8. 2026Do Thinking Tokens Help with Safety?cited by 3
  9. 2026Context-CoT: Enhancing Context Learning via High-Quality Reasoning Synthesiscited by 2
  10. 2026Scaling Large Reasoning Models beyond Human Supervision: A Path toward Superintelligencecited by 2
  11. 2026Vision-language models lag human performance on physical dynamics and intent reasoningcited by 1
  12. 2026Alignment Tuning for Large Language Models: A Data-Centric Lens on Alignment Data Pipelinescited by 1
  13. 2026When the Strongest Teacher Is Not the Best Teacher: Student-Centric Answer Selectioncited by 1
  14. 2026FAQ: Mitigating Quantization Error via Regenerating Calibration Data with Family-Aware Quantizationcited by 0
  15. 2026Cognitively Layered Data Synthesis for Domain Adaptation of LLMs to Space Situational Awarenesscited by 0
  16. 2026Eliciting Medical Reasoning with Knowledge-enhanced Data Synthesis: A Semi-Supervised Reinforcement Learning Approachcited by 0
  17. 2026Knowledge-to-Verification: Exploring RLVR for LLMs in Knowledge-Intensive Domainscited by 0
  18. 2026DataFoundry: Evolving Data Preparators via Recursive Self-Improvementcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025CoT-based Synthesizer: Enhancing LLM Performance through Answer Synthesis
  2. 2025The Challenge of Teaching Reasoning to LLMs Without RL or Distillation
  3. 2025Code Execution as Grounded Supervision for LLM Reasoning
  4. 2026Evaluating Prompting and Execution-Based Methods for Deterministic Computation in LLMs
  5. 2025Zero-Shot Verification-guided Chain of Thoughts
  6. 2022Large Language Models are Better Reasoners with Self-Verification
  7. 2025LLMs Can Easily Learn to Reason from Demonstrations Structure, not content, is what matters!
  8. 2025When Thinking Fails: The Pitfalls of Reasoning for Instruction-Following in LLMs