mlsanityData to 7 Oct 2026

Leveraging Grammar and Reinforcement Learning for Neural Program Synthesis

arXiv 1805.04276 ↗First posted 11 May 2018cs.LG stat.MLcited by 83

Not on the map. Why?

Abstract (arXiv)

Program synthesis is the task of automatically generating a program consistent with a specification. Recent years have seen proposal of a number of neural approaches for program synthesis, many of which adopt a sequence generation paradigm similar to neural machine translation, in which sequence-to-sequence models are trained to maximize the likelihood of known reference programs. While achieving impressive results, this strategy has two key limitations. First, it ignores Program Aliasing: the fact that many different programs may satisfy a given specification (especially with incomplete specifications such as a few input-output examples). By maximizing the likelihood of only a single reference program, it penalizes many semantically correct programs, which can adversely affect the synthesizer performance. Second, this strategy overlooks the fact that programs have a strict syntax that can be efficiently checked. To address the first limitation, we perform reinforcement learning on top of a supervised model with an objective that explicitly maximizes the likelihood of generating semantically correct programs. For addressing the second limitation, we introduce a training procedure that directly maximizes the probability of generating syntactically correct programs that fulfill the specification. We show that our contributions lead to improved accuracy of the models, especially in cases where the training data is limited.

Key facts. Cited by 83 papers in the map (data through 7 Oct 2026). First posted on arXiv 11 May 2018.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2019: 2; 2020: 0; 2021: 17; 2022: 15; 2023: 24; 2024: 16; 2025: 3; 2026: 6.

83 papers.

Sort:

By area:

  1. 2023Teaching Large Language Models to Self-Debugcited by 963
  2. 2022CodeRL: Mastering Code Generation through Pretrained Models and Deep Reinforcement Learningcited by 342
  3. 2023InterCode: Standardizing and Benchmarking Interactive Coding with Execution Feedbackcited by 229
  4. 2023Planning with Large Language Models for Code Generationcited by 179
  5. 2023Learning Performance-Improving Code Editscited by 103
  6. 2021Neural Production Systems: Learning Rule-Governed Visual Dynamicscited by 75
  7. 2021Learning to Synthesize Programs as Interpretable and Generalizable Policiescited by 59
  8. 2021From LSAT: The Progress and Challenges of Complex Reasoningcited by 43
  9. 2021SpreadsheetCoder: Formula Prediction from Semi-structured Contextcited by 41
  10. 2022Learning Math Reasoning from Self-Sampled Correct and Partially-Correct Solutionscited by 38
  11. 2021Latent Execution for Neural Program Synthesiscited by 35
  12. 2024PolyNet: Learning Diverse Solution Strategies for Neural Combinatorial Optimizationcited by 35
38 more
  1. 2019Few-Shot Bayesian Imitation Learning with Logical Program Policiescited by 30
  2. 2026Maximum Likelihood Reinforcement Learningcited by 29
  3. 2023Emergent Representations of Program Semantics in Language Models Trained on Programscited by 28
  4. 2022How to Reuse and Compose Knowledge for a Lifetime of Tasks: A Survey on Continual Learning and Functional Compositioncited by 27
  5. 2023Outline, Then Details: Syntactically Guided Coarse-To-Fine Code Generationcited by 27
  6. 2024CodeIt: Self-Improving Language Models with Prioritized Hindsight Replaycited by 23
  7. 2023Hierarchical Programmatic Reinforcement Learning via Learning to Compose Programscited by 20
  8. 2021ProTo: Program-Guided Transformer for Program-Guided Taskscited by 19
  9. 2023Proximal Curriculum for Reinforcement Learning Agentscited by 19
  10. 2025Hybrid Reinforcement: When Reward Is Sparse, It's Better to Be Densecited by 19
  11. 2022CrossBeam: Learning to Search in Bottom-Up Program Synthesiscited by 18
  12. 2023ExeDec: Execution Decomposition for Compositional Generalization in Neural Program Synthesiscited by 17
  13. 2022GALOIS: Boosting Deep Reinforcement Learning via Generalizable Logic Synthesiscited by 16
  14. 2023Neural Task Synthesis for Visual Programmingcited by 13
  15. 2024Synthesizing Programmatic Reinforcement Learning Policies with Large Language Model Guided Searchcited by 13
  16. 2023Adaptive Scaffolding in Block-Based Programming via Synthesizing New Tasks as Pop Quizzescited by 10
  17. 2024Guiding Enumerative Program Synthesis with Large Language Modelscited by 9
  18. 2021DeepQAMVS: Query-Aware Hierarchical Pointer Networks for Multi-Video Summarizationcited by 8
  19. 2021Learning to Combine Per-Example Solutions for Neural Program Synthesiscited by 8
  20. 2021Interpretable Model-based Hierarchical Reinforcement Learning using Inductive Logic Programmingcited by 8
  21. 2022Synthesizing explainable counterfactual policies for algorithmic recourse with program synthesiscited by 8
  22. 2022From {Solution Synthesis} to {Student Attempt Synthesis} for Block-Based Visual Programming Taskscited by 8
  23. 2021Program Synthesis Guided Reinforcement Learning for Partially Observed Environmentscited by 7
  24. 2021Transcribing Natural Languages for The Deaf via Neural Editing Programscited by 7
  25. 2023Hierarchical Neural Program Synthesiscited by 7
  26. 2023Program Synthesis with Best-First Bottom-Up Searchcited by 7
  27. 2024Grounding Data Science Code Generation with Input-Output Specificationscited by 7
  28. 2024Hints-In-Browser: Benchmarking Language Models for Programming Feedback Generationcited by 7
  29. 2021Scaling Neural Program Synthesis with Distribution-based Searchcited by 6
  30. 2022Compositional Generalization and Decomposition in Neural Program Synthesiscited by 6
  31. 2022E-MAPP: Efficient Multi-Agent Reinforcement Learning with Parallel Program Guidancecited by 6
  32. 2024Procedural Synthesis of Synthesizable Moleculescited by 6
  33. 2019Hierarchical Variational Imitation Learning of Control Programscited by 5
  34. 2022Learning to Find Proofs and Theorems by Learning to Refine Search Strategies: The Case of Loop Invariant Synthesiscited by 5
  35. 2022Learning to Improve Code Efficiencycited by 4
  36. 2024Learning big logical rules by joining small rulescited by 4
  37. 2024Towards Generalizable Agents in Text-Based Educational Environments: A Study of Integrating RL with LLMscited by 4
  38. 2024Program Synthesis Benchmark for Visual Programming in XLogoOnline Environmentcited by 4

50 most cited of 83 All 83 (JSON)