Local Success Does Not Compose: Benchmarking Large Language Models for Compositional Formal Verification
Year: 2025
Type: preprint
Abstract: We introduce DafnyCOMP, a benchmark for evaluating large language models (LLMs) on compositional specification generation in Dafny. Unlike prior benchmarks that focus on single-function tasks, DafnyCOMP targets programs composed of multiple interacting functions with data dependencies, requiring reasoning across component boundaries. The benchmark consists of 300 automatically synthesized multi-function programs. We evaluate several state-of-the-art LLM families and find that, while they perform well on single-function verification, their performance drops sharply on compositional tasks. Analysis reveals systematic failures in cross-functional reasoning, including fragile specifications, misalignment between implementations and proofs, and unstable reasoning. DafnyCOMP thus provides a diagnostic tool for measuring progress toward reliable, verifiable, and compositional code generation with LLMs.
Source: arXiv (Cornell University)
Authors: Xu, Xu, Li, Xin, Qu, Xingwei, Fu, Jie, Yuan, Binhang
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7083833859
Cites:
Cited by:
Related to: