Abstract Syntax Networks for Code Generation and Semantic Parsing

2017/04/25 by Rabinovich, Maxim, Stern, Mitchell, Klein, Dan · 7 citations
#Artificial Intelligence (cs.AI) #Computation and Language (cs.CL) #FOS: Computer and information sciences #Machine Learning (cs.LG) #Machine Learning (stat.ML)

paper · doi:10.48550/arxiv.1704.07535

Abstract

Tasks like code generation and semantic parsing require mapping unstructured (or partially structured) inputs to well-formed, executable outputs. We introduce abstract syntax networks, a modeling framework for these problems. The outputs are represented as abstract syntax trees (ASTs) and constructed by a decoder with a dynamically-determined modular structure paralleling the structure of the output tree. On the benchmark Hearthstone dataset for code generation, our model obtains 79.2 BLEU and 22.7% exact match accuracy, compared to previous state-of-the-art values of 67.1 and 6.1%. Furthermore, we perform competitively on the Atis, Jobs, and Geo semantic parsing datasets with no task-specific engineering.

Cited by

• Pointer-Augmented Autoregressive Generation of Patent Claims with Joint Topology and Content Decoding
• GramTrans: A Better Code Representation Approach in Code Generation
• Understanding Benchmark Language Under Weakened Formal Semantics
• Automating Code Generation for Semiconductor Equipment Control from Developer Utterances with LLMs
• Symbolic Feedforward Networks for Probabilistic Finite Automata: Exact Simulation and Learnability
• TreeDiff: AST-Guided Code Generation with Diffusion LLMs
• CrossPL: Evaluating Large Language Models on Cross Programming Language Code Generation

Related

• A Syntactic Neural Model for General-Purpose Code Generation
• GN-Transformer: Fusing Sequence and Graph Representation for Improved Code Summarization
• Abstract Syntax Tree for Programming Language Understanding and Representation: How Far Are We?
• GNN-Coder: Boosting Semantic Code Retrieval with Combined GNNs and Transformer
• The ModelCC Model-Based Parser Generator