Open AccessProceedings Article
Execution-Guided Neural Program Synthesis.
Xinyun Chen,Chang Liu,Dawn Song +2 more
- 27 Sep 2018
TL;DR: This work proposes two simple yet principled techniques to better leverage the semantic information, which are execution-guided synthesis and synthesizer ensemble that are general enough to be combined with any existing encoder-decoder-style neural program synthesizer.
read more
Abstract: Neural program synthesis from input-output examples has attracted an increasing interest from both the machine learning and the programming language community. Most existing neural program synthesis approaches employ an encoder-decoder architecture, which uses an encoder to compute the embedding of the given inputoutput examples, as well as a decoder to generate the program from the embedding following a given syntax. Although such approaches achieve a reasonable performance on simple tasks such as FlashFill, on more complex tasks such as Karel, the state-of-the-art approach can only achieve an accuracy of around 77%. We observe that the main drawback of existing approaches is that the semantic information is greatly under-utilized. In this work, we propose two simple yet principled techniques to better leverage the semantic information, which are execution-guided synthesis and synthesizer ensemble. These techniques are general enough to be combined with any existing encoder-decoder-style neural program synthesizer. Applying our techniques to the Karel dataset, we can boost the accuracy from around 77% to more than 90%.
read more
Chat with Paper
AI Agents for this Paper
Find similar papers on Google Scholar, PubMed and Arxiv
Write a critical review of this paper
Analyze citations of this paper to find unaddressed research gaps
Citations
Voyager: An Open-Ended Embodied Agent with Large Language Models
Guanzhi Wang,Yunfan Jiang,Ajay Mandlekar,Chaowei Xiao,Yuke Zhu,Linxi Fan,Animashree Anandkumar +6 more
TL;DR: Voyager as discussed by the authors is the first LLM-powered embodied lifelong learning agent in Minecraft that continuously explores the world, acquires diverse skills, and makes novel discoveries without human intervention, and it is able to utilize the learned skill library in a new Minecraft world to solve novel tasks from scratch.
410
•Book
Synthetic Data for Deep Learning
Sergey I. Nikolenko
- 26 Jun 2021
TL;DR: The synthetic-to-real domain adaptation problem that inevitably arises in applications of synthetic data is discussed, including synthetic- to-real refinement with GAN-based models and domain adaptation at the feature/model level without explicit data transformations.
391
Teaching Large Language Models to Self-Debug
TL;DR: Self-Debugging as discussed by the authors proposes to train a large language model to debug its predicted program via few-shot demonstrations, i.e., without any feedback on the code correctness or error messages, the model can identify its mistakes by explaining the generated code in natural language.
324
CodeT: Code Generation with Generated Tests
Fengji Zhang,Anh Nguyen,Daoguang Zan,Zeqi Lin,Jian-Guang Lou,Weizhu Chen +5 more
- 21 Jul 2022
TL;DR: This paper explores the use of pre-trained language models to automatically generate test cases, calling it CODET: CODE generation with generated Tests, and then chooses the best solution based on a dual execution agreement with both the generated test cases and other generated solutions.
LEVER: Learning to Verify Language-to-Code Generation with Execution
Ansong Ni,Srinivasan Iyer,Dragomir R. Radev,Veselin Stoyanov,Wen-tau Yih,Sida Wang,Xi Victoria Lin +6 more
TL;DR: LeVER as discussed by the authors learns to verify the generated programs with their execution results by training verifiers to determine whether a program sampled from the LLMs is correct or not based on the natural language input, the program itself and its execution results.
References
•Proceedings Article
Adam: A Method for Stochastic Optimization
Diederik P. Kingma,Jimmy Ba +1 more
- 01 Jan 2015
TL;DR: This work introduces Adam, an algorithm for first-order gradient-based optimization of stochastic objective functions, based on adaptive estimates of lower-order moments, and provides a regret bound on the convergence rate that is comparable to the best known results under the online convex optimization framework.
138.5K
Simple Statistical Gradient-Following Algorithms for Connectionist Reinforcement Learning
TL;DR: This article presents a general class of associative reinforcement learning algorithms for connectionist networks containing stochastic units that are shown to make weight adjustments in a direction that lies along the gradient of expected reinforcement in both immediate-reinforcement tasks and certain limited forms of delayed-reInforcement tasks, and they do this without explicitly computing gradient estimates.
Automating string processing in spreadsheets using input-output examples
Sumit Gulwani
- 26 Jan 2011
TL;DR: The design of a string programming/expression language that supports restricted forms of regular expressions, conditionals and loops is described and an algorithm based on several novel concepts for synthesizing a desired program in this language is described from input-output examples.
•Posted Content
SQLNet: Generating Structured Queries From Natural Language Without Reinforcement Learning
Xiaojun Xu,Chang Liu,Dawn Song +2 more
TL;DR: A sketch-based approach where the sketch contains a dependency graph, so that one prediction can be done by taking into consideration only the previous predictions that it depends on, and it is shown that SQLNet can outperform the prior art by 9% to 13% on the WikiSQL task.
•Proceedings Article
Inferring algorithmic patterns with stack-augmented recurrent nets
Armand Joulin,Tomas Mikolov +1 more
- 07 Dec 2015
TL;DR: The limitations of standard deep learning approaches are discussed and it is shown that some of these limitations can be overcome by learning how to grow the complexity of a model in a structured way.