CITATION.cff

cff-version: 1.2.0
message: "If you use this software, please cite the ACL 2026 Findings paper below."
title: "SynthFix: Adaptive Neuro-Symbolic Code Vulnerability Repair"
abstract: >
  SynthFix is a training library that combines supervised fine-tuning
  with a reward-driven update through a small neural router that gates
  the reward signal per sample. The reward is a split symbolic score
  over AST, control-flow, semantic, and surface-similarity
  components, and can also be used as features for an inference-time
  best-of-K reranker.
authors:
  - family-names: "SynthFix"
    given-names: "Authors"
version: "0.1.0"
date-released: 2026-04-18
license: "See LICENSE"
preferred-citation:
  type: conference-paper
  title: "SynthFix: Adaptive Neuro-Symbolic Code Vulnerability Repair"
  authors:
    - family-names: "SynthFix"
      given-names: "Authors"
  collection-title: "Findings of the Association for Computational Linguistics: ACL 2026"
  year: 2026