mlsanityData to 14 Oct 2026

Language Models Can Teach Themselves to Program Better

arXiv 2207.14502 ↗First posted 29 Jul 2022cs.LG cs.AIcited by 78

Not on the map. Why?

Cited by 20 in Synthetic data and self-improvement for LLMs, but placed by its own abstract, which didn’t fit.

Abstract (arXiv)

Recent Language Models (LMs) achieve breakthrough performance in code generation when trained on human-authored problems, even solving some competitive-programming problems. Self-play has proven useful in games such as Go, and thus it is natural to ask whether LMs can generate their own instructive programming problems to improve their performance. We show that it is possible for an LM to synthesize programming problems and solutions, which are filtered for correctness by a Python interpreter. The LM's performance is then seen to improve when it is fine-tuned on its own synthetic problems and verified solutions; thus the model 'improves itself' using the Python interpreter. Problems are specified formally as programming puzzles [Schuster et al., 2021], a code-based problem format where solutions can easily be verified for correctness by execution. In experiments on publicly-available LMs, test accuracy more than doubles. This work demonstrates the potential for code LMs, with an interpreter, to generate instructive problems and improve their own performance.

Key facts. Cited by 78 papers in the map (data through 14 Oct 2026). Builds on 28 of its 49 references that are in the map. First posted on arXiv 29 Jul 2022.

Builds on

28 of 49 references.

References · 28

  1. 2014Adam: A Method for Stochastic Optimizationcited by 37,240
  2. 2020Language Models are Few-Shot Learnerscited by 32,419
  3. 2015Distilling the Knowledge in a Neural Networkcited by 10,324
  4. 2021Training Verifiers to Solve Math Word Problemscited by 10,109
  5. 2021Evaluating Large Language Models Trained on Codecited by 8,320
  6. 2022PaLM: Scaling Language Modeling with Pathwayscited by 5,326
  7. 2021Program Synthesis with Large Language Modelscited by 3,637
  8. 2022Training Compute-Optimal Large Language Modelscited by 3,019
  9. 2018UMAP: Uniform Manifold Approximation and Projection for Dimension Reductioncited by 2,898
  10. 2020The Pile: An 800GB Dataset of Diverse Text for Language Modelingcited by 2,411
  11. 2017Decoupled Weight Decay Regularizationcited by 1,831
  12. 2020Knowledge Distillation: A Surveycited by 1,603
16 more
  1. 2022Competition-Level Code Generation with AlphaCodecited by 1,462
  2. 2021Measuring Coding Challenge Competence With APPScited by 902
  3. 2020Generative Language Modeling for Automated Theorem Provingcited by 388
  4. 2022CodeRL: Mastering Code Generation through Pretrained Models and Deep Reinforcement Learningcited by 358
  5. 2021Symbolic Knowledge Distillation: from General Language Models to Commonsense Modelscited by 270
  6. 2022Generating Training Data with Language Models: Towards Zero-Shot Language Understandingcited by 199
  7. 2021Generating Datasets with Pretrained Language Modelscited by 193
  8. 2016DeepCoder: Learning to Write Programscited by 149
  9. 2022Automatic Generation of Programming Exercises and Code Explanations using Large Language Modelscited by 126
  10. 2019Write, Execute, Assess: Program Synthesis with a REPLcited by 93
  11. 2021Leveraging Automated Unit Tests for Unsupervised Code Translationcited by 82
  12. 2021LIME: Learning Inductive Bias for Primitives of Mathematical Reasoningcited by 46
  13. 2021Programming Puzzlescited by 25
  14. 2019Synthetic Datasets for Neural Program Synthesiscited by 15
  15. 2018NAPS: Natural Program Synthesis Datasetcited by 14
  16. 2020Learning to Prove from Synthetic Theoremscited by 13

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2022: 3; 2023: 16; 2024: 33; 2025: 19; 2026: 7.

78 papers.

Sort:

By area:

  1. 2024Agent Workflow Memorycited by 383
  2. 2024Quiet-STaR: Language Models Can Teach Themselves to Think Before Speakingcited by 301
  3. 2024CRUXEval: A Benchmark for Code Reasoning, Understanding and Executioncited by 300
  4. 2025SWE-smith: Scaling Data for Software Engineering Agentscited by 285
  5. 2024Two Tales of Persona in LLMs: A Survey of Role-Playing and Personalizationcited by 208
  6. 2022Will we run out of data? Limits of LLM scaling based on human-generated datacited by 195
  7. 2023Self-Taught Optimizer (STOP): Recursively Self-Improving Code Generationcited by 148
  8. 2024LLM-SR: Scientific Equation Discovery via Programming with Large Language Modelscited by 133
  9. 2024Meta-Prompting: Enhancing Language Models with Task-Agnostic Scaffoldingcited by 117
  10. 2023Learning Performance-Improving Code Editscited by 106
  11. 2024Best Practices and Lessons Learned on Synthetic Datacited by 106
  12. 2022Execution-Based Evaluation for Open-Domain Code Generationcited by 91
38 more
  1. 2024Buffer of Thoughts: Thought-Augmented Reasoning with Large Language Modelscited by 91
  2. 2023Phenomenal Yet Puzzling: Testing Inductive Reasoning Capabilities of Language Models with Hypothesis Refinementcited by 89
  3. 2024If LLM Is the Wizard, Then Code Is the Wand: A Survey on How Code Empowers Large Language Models to Serve as Intelligent Agentscited by 79
  4. 2025STP: Self-play LLM Theorem Provers with Iterative Conjecturing and Provingcited by 79
  5. 2024LLM2LLM: Boosting LLMs with Novel Iterative Data Enhancementcited by 70
  6. 2024SelfCodeAlign: Self-Alignment for Code Generationcited by 62
  7. 2025Synthetic Data Generation Using Large Language Models: Advances in Text and Codecited by 61
  8. 2023LILO: Learning Interpretable Libraries by Compressing and Documenting Codecited by 49
  9. 2024Puzzle Solving using Reasoning of Large Language Models: A Surveycited by 45
  10. 2024Beyond Model Collapse: Scaling Up with Synthesized Data Requires Verificationcited by 45
  11. 2024Learning Formal Mathematics From Intrinsic Motivationcited by 45
  12. 2023SelfEvolve: A Code Evolution Framework via Large Language Modelscited by 43
  13. 2025AlphaAgent: LLM-Driven Alpha Mining with Regularized Exploration to Counteract Alpha Decaycited by 42
  14. 2024A Survey on Data Synthesis and Augmentation for Large Language Modelscited by 41
  15. 2025Large Language Models for Planning: A Comprehensive and Systematic Surveycited by 40
  16. 2025LLM-FE: Automated Feature Engineering for Tabular Data with LLMs as Evolutionary Optimizerscited by 38
  17. 2024QuantAgent: Seeking Holy Grail in Trading by Self-Improving Large Language Modelcited by 30
  18. 2023AI capabilities can be significantly improved without expensive retrainingcited by 28
  19. 2023LLM-Assisted Code Cleaning For Training Accurate Code Generatorscited by 27
  20. 2024DARG: Dynamic Evaluation of Large Language Models via Adaptive Reasoning Graphcited by 26
  21. 2023LeTI: Learning to Generate from Textual Interactionscited by 25
  22. 2024CodeIt: Self-Improving Language Models with Prioritized Hindsight Replaycited by 23
  23. 2023Language models are weak learnerscited by 22
  24. 2022I2D2: Inductive Knowledge Distillation with NeuroLogic and Self-Imitationcited by 21
  25. 2024Bootstrapping Language-Guided Navigation Learning with Self-Refining Data Flywheelcited by 20
  26. 2023Incremental Learning of Humanoid Robot Behavior from Natural Interaction and Large Language Modelscited by 19
  27. 2024SPARQL Generation: an analysis on fine-tuning OpenLLaMA for Question Answering over a Life Science Knowledge Graphcited by 19
  28. 2025Beyond Prompt Content: Enhancing LLM Performance via Content-Format Integrated Prompt Optimizationcited by 16
  29. 2024Mitigating Tail Narrowing in LLM Self-Improvement via Socratic-Guided Samplingcited by 13
  30. 2025Propose, Solve, Verify: Self-Play Through Formal Verificationcited by 13
  31. 2023Generating Data for Symbolic Language with Large Language Modelscited by 12
  32. 2024Self-Training Large Language Models for Improved Visual Program Synthesis With Visual Reinforcementcited by 12
  33. 2024Data for Mathematical Copilots: Better Ways of Presenting Proofs for Machine Learningcited by 11
  34. 2024PLUM: Improving Code LMs with Execution-Guided On-Policy Preference Learning Driven By Synthetic Test Casescited by 8
  35. 2025Why Less is More (Sometimes): A Theory of Data Curationcited by 8
  36. 2024Fictitious Synthetic Data Can Improve LLM Factuality via Prerequisite Learningcited by 7
  37. 2023ACES: Generating Diverse Programming Puzzles with with Autotelic Generative Modelscited by 6
  38. 2024SELU: Self-Learning Embodied MLLMs in Unknown Environmentscited by 6

50 most cited of 78 All 78 (JSON)