Skip to content

codimango/fvspec

 
 

Repository files navigation

fvspec

Lean benchmark for real-world SWE tasks.

We scrape github for PBTs in hypothesis and use an LLM to transpile them to Lean specs, autoformalizing each function under test along the way.

Repo

./leaderboard

The website

./benchmark

The benchmark generation pipeline and postproduction tools. Includes unified formalization agent (produces both Impl.lean and Spec.lean), quality assessment, and post-processing (turncount, merge, metrics, grading). Requires ANTHROPIC_API_KEY in .env. Source PBTs are pulled automatically from the GaloisInc/fvspec-pbt dataset on the Hugging Face Hub; to run against a local file instead, drop a JSONL under ./benchmark/data and pass its name via --datafile.

./baselines

Conducting performance baselines on the benchmark

Dependencies

All you need are uv and elan to work with the pipelines or evals, but npm is required for the website. Tip: run lake build in ./benchmark/lake-template once to prepopulate the .lake dir for performance.

Acknowledgements

This project is funded by the Advanced Research + Invention Agency (ARIA).

License

This work is made available under both an MIT license and an Apache 2 license such that users can decide which to utilize.

About

Benchmark suite for helping evaluate how AIs perform on formal verification related tasks

Resources

Stars

Watchers

Forks

Releases

Packages

Used by

Contributors

Languages