This method is designed to overcome challenges in a standard autoregressive approach, allowing the model to make targeted, grammatically correct changesThis method is designed to overcome challenges in a standard autoregressive approach, allowing the model to make targeted, grammatically correct changes

Syntactically Valid Code Editing: A Training Methodology for Neural Program Synthesis

Par : Hackernoon

2025/09/25 00:00

Table of Links

Abstract and 1. Introduction

Background & Related Work
Method

3.1 Sampling Small Mutations

3.2 Policy

3.3 Value Network & Search

3.4 Architecture
Experiments

4.1 Environments

4.2 Baselines

4.3 Ablations
Conclusion, Acknowledgments and Disclosure of Funding, and References

\

Appendix

A. Mutation Algorithm

B. Context-Free Grammars

C. Sketch Simulation

D. Complexity Filtering

E. Tree Path Algorithm

F. Implementation Details

3.4 Architecture

\ We add two additional types of tokens: an token, which serves as a start-of-sentence token for the model; and tokens, which allow the model to reference positions within its context. Given a current image, a target image, and a current tokenized program, we train this transformer model to predict the edit position and the replacement text autoregressively. While making predictions, the decoding is constrained under the grammar. We mask out the prediction logits to only include edit positions that represent nodes in the syntax tree, and only produce replacements that are syntactically valid for the selected edit position.

\ We set σsmall = 2, which means the network is only allowed to produce edits with fewer than two primitives. For training data, we sample an infinite stream of random expressions from the CFG. We choose a random number of noise steps, s ∈ [1, 5], to produce a mutated expression. For some percentage of the examples, ρ, we instead sample a completely random new expression as our mutated expression. We trained for 3 days for the environments we tested on a single Nvidia A6000 GPU.

:::info Authors:

(1) Shreyas Kapur, University of California, Berkeley (srkp@cs.berkeley.edu);

(2) Erik Jenner, University of California, Berkeley (jenner@cs.berkeley.edu);

(3) Stuart Russell, University of California, Berkeley (russell@cs.berkeley.edu).

:::

:::info This paper is available on arxiv under CC BY-SA 4.0 DEED license.

:::

Clause de non-responsabilité : les articles republiés sur ce site proviennent de plateformes publiques et sont fournis à titre informatif uniquement. Ils ne reflètent pas nécessairement les opinions de MEXC. Tous les droits restent la propriété des auteurs d'origine. Si vous estimez qu'un contenu porte atteinte aux droits d'un tiers, veuillez contacter service@support.mexc.com pour demander sa suppression. MEXC ne garantit ni l'exactitude, ni l'exhaustivité, ni l'actualité des contenus, et décline toute responsabilité quant aux actions entreprises sur la base des informations fournies. Ces contenus ne constituent pas des conseils financiers, juridiques ou professionnels, et ne doivent pas être interprétés comme une recommandation ou une approbation de la part de MEXC.

Partager des idées