0
views
0
recommends
+1 Recommend
0 collections
    0
    shares
      • Record: found
      • Abstract: found
      • Article: found
      Is Open Access

      Can LLMs Enable Verification in Mainstream Programming?

      Preprint

      Read this article at

      Bookmark
          There is no author summary for this article yet. Authors can add summaries to their articles on ScienceOpen to make them more accessible to a non-specialist audience.

          Abstract

          Although formal methods are capable of producing reliable software, they have seen minimal adoption in everyday programming. Automatic code generation using large language models is becoming increasingly widespread, but it rarely considers producing strong correctness guarantees. In this study, we explore the ability of LLMs to produce verified code in three verification languages (Dafny, Nagini, and Verus). To do so, we use manually curated datasets derived from the state-ofthe-art Python benchmark, HumanEval. We also assess what types of information are sufficient to achieve good-quality results.

          Related collections

          Author and article information

          Journal
          18 March 2025
          Article
          2503.14183
          d4792f34-adad-4dfb-b228-ef02ad56b8bc

          http://creativecommons.org/licenses/by/4.0/

          History
          Custom metadata
          cs.SE cs.AI cs.PL

          Software engineering,Programming languages,Artificial intelligence
          Software engineering, Programming languages, Artificial intelligence

          Comments

          Comment on this article