neural program synthesis
play

Neural Program Synthesis Rishabh Singh, Google Brain Great - PowerPoint PPT Presentation

Neural Program Synthesis Rishabh Singh, Google Brain Great Collaborators! Deep Learning and Evolutionary Progression Vision Speech Language Deep Learning and Evolutionary Progression Vision Speech Language Programming Deep Learning


  1. Reverse-Recursive Input: root representation from recursive pass

  2. Reverse-Recursive Input: root representation from recursive pass

  3. Reverse-Recursive Input: root representation from recursive pass Output: Global leaf representations.

  4. Reverse-Recursive Input: root representation from recursive pass Output: Global leaf representations.

  5. Reverse-Recursive Input: root representation from recursive pass Output: Global leaf representations.

  6. Cross-Correlation I/O Encoder

  7. Cross-Correlation I/O Encoder

  8. Cross-Correlation I/O Encoder

  9. Synthetic Data Results (< 13 AST)

  10. FlashFill Benchmarks

  11. FlashFill Benchmarks Batching Trees for larger programs

  12. FlashFill Benchmarks Batching Trees for larger programs R3NN for contextual program embeddings

  13. RobustFill [ICML 2017] J. Devlin, J. Uesato, S. Bhuptiraju, R. Singh, A. Mohamed, P . Kohli

  14. RobustFill [ICML 2017] J. Devlin, J. Uesato, S. Bhuptiraju, R. Singh, A. Mohamed, P . Kohli

  15. RobustFill [ICML 2017] J. Devlin, J. Uesato, S. Bhuptiraju, R. Singh, A. Mohamed, P . Kohli

  16. RobustFill [ICML 2017] J. Devlin, J. Uesato, S. Bhuptiraju, R. Singh, A. Mohamed, P . Kohli

  17. Multiple I/O Examples

  18. Multiple I/O Examples

  19. Multiple I/O Examples

  20. Extended DSL

  21. 92% Generalization Accuracy

  22. Robustness with Noise

  23. Incorrect Generalization

Recommend


More recommend