mlsanityData to 14 Oct 2026

Neural Programmer-Interpreters

arXiv 1511.06279 ↗First posted 19 Nov 2015cs.LG cs.NEcited by 101

Not on the map. Why?

Abstract (arXiv)

We propose the neural programmer-interpreter (NPI): a recurrent and compositional neural network that learns to represent and execute programs. NPI has three learnable components: a task-agnostic recurrent core, a persistent key-value program memory, and domain-specific encoders that enable a single NPI to operate in multiple perceptually diverse environments with distinct affordances. By learning to compose lower-level programs to express higher-level programs, NPI reduces sample complexity and increases generalization ability compared to sequence-to-sequence LSTMs. The program memory allows efficient learning of additional tasks by building on existing programs. NPI can also harness the environment (e.g. a scratch pad with read-write pointers) to cache intermediate results of computation, lessening the long-term memory burden on recurrent hidden units. In this work we train the NPI with fully-supervised execution traces; each program has example sequences of calls to the immediate subprograms conditioned on the input. Rather than training on a huge number of relatively weak labels, NPI learns from a small number of rich examples. We demonstrate the capability of our model to learn several types of compositional programs: addition, sorting, and canonicalizing 3D models. Furthermore, a single NPI learns to execute these programs and all 21 associated subprograms.

Key facts. Cited by 101 papers in the map (data through 14 Oct 2026). First posted on arXiv 19 Nov 2015.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2019: 1; 2020: 0; 2021: 32; 2022: 13; 2023: 20; 2024: 7; 2025: 10; 2026: 18.

101 papers.

Sort:

By area:

  1. 2021Evaluating Large Language Models Trained on Codecited by 8,320
  2. 2021Geometric Deep Learning: Grids, Groups, Graphs, Geodesics, and Gaugescited by 1,167
  3. 2022Solving math word problems with process- and outcome-based feedbackcited by 901
  4. 2021Show Your Work: Scratchpads for Intermediate Computation with Language Modelscited by 886
  5. 2022A Generalist Agentcited by 808
  6. 2021CodeXGLUE: A Machine Learning Benchmark Dataset for Code Understanding and Generationcited by 744
  7. 2022Grokking: Generalization Beyond Overfitting on Small Algorithmic Datasetscited by 629
  8. 2023WizardMath: Empowering Mathematical Reasoning for Large Language Models via Reinforced Evol-Instructcited by 625
  9. 2022Decomposed Prompting: A Modular Approach for Solving Complex Taskscited by 558
  10. 2021Combinatorial optimization and reasoning with graph neural networkscited by 268
  11. 2022LIFT: Language-Interfaced Fine-Tuning for Non-Language Machine Learning Taskscited by 148
  12. 2021Neural Algorithmic Reasoningcited by 94
38 more
  1. 2021Coordination Among Neural Modules Through a Shared Global Workspacecited by 85
  2. 2023Modular Deep Learningcited by 83
  3. 2021Neural Production Systems: Learning Rule-Governed Visual Dynamicscited by 75
  4. 2023Large Language Models can Learn Rulescited by 69
  5. 2021Learning to Synthesize Programs as Interpretable and Generalizable Policiescited by 59
  6. 2024Combining Induction and Transduction for Abstract Reasoningcited by 53
  7. 2021Evolving Reinforcement Learning Algorithmscited by 49
  8. 2021Inductive logic programming at 30cited by 47
  9. 2023Learning Transformer Programscited by 41
  10. 2021Discrete-Valued Neural Communicationcited by 39
  11. 2021Fast and Slow Learning of Recurrent Independent Mechanismscited by 36
  12. 2024Learning Iterative Reasoning through Energy Diffusioncited by 36
  13. 2023Code Execution with Pre-trained Language Modelscited by 34
  14. 2022Learning Iterative Reasoning through Energy Minimizationcited by 32
  15. 2022Chain of Thought Imitation with Procedure Cloningcited by 30
  16. 2024Simulation of Graph Algorithms with Looped Transformerscited by 30
  17. 2026DeltaEvolve: Accelerating Scientific Discovery through Momentum-Driven Evolutioncited by 26
  18. 2023Recursion of Thought: A Divide-and-Conquer Approach to Multi-Context Reasoning with Language Modelscited by 24
  19. 2021How to transfer algorithmic reasoning knowledge to learn new algorithms?cited by 22
  20. 2021ProTo: Program-Guided Transformer for Program-Guided Taskscited by 21
  21. 2023Dual Algorithmic Reasoningcited by 21
  22. 2023Hierarchical Programmatic Reinforcement Learning via Learning to Compose Programscited by 20
  23. 2021Transformers with Competitive Ensembles of Independent Mechanismscited by 19
  24. 2021Pushing the Limits of Rule Reasoning in Transformers through Natural Language Satisfiabilitycited by 18
  25. 2023Chain-of-Thought Predictive Controlcited by 18
  26. 2022Neural Approximation of Graph Topological Featurescited by 17
  27. 2023Neural Relational Inference with Fast Modular Meta-learningcited by 16
  28. 2022Multi-Step Deductive Reasoning Over Natural Language: An Empirical Study on Out-of-Distribution Generalisationcited by 15
  29. 2021UnweaveNet: Unweaving Activity Storiescited by 14
  30. 2022Static Prediction of Runtime Errors by Learning to Execute Programs with External Resource Descriptionscited by 11
  31. 2025Image Editing As Programs with Diffusion Modelscited by 10
  32. 2021pix2rule: End-to-end Neuro-symbolic Rule Learningcited by 9
  33. 2023Recursive Algorithmic Reasoningcited by 9
  34. 2024Discrete Neural Algorithmic Reasoningcited by 9
  35. 2021Recognizing and Verifying Mathematical Equations using Multiplicative Differential Neural Unitscited by 8
  36. 2021Transcribing Natural Languages for The Deaf via Neural Editing Programscited by 7
  37. 2023Hierarchical Neural Program Synthesiscited by 7
  38. 2026Tiny Autoregressive Recursive Modelscited by 6

50 most cited of 101 All 101 (JSON)