1 paper
Sergiu Bursuc, Theodore Ehrenborg, Shaowei Lin +10
We present and test the largest benchmark for vericoding, LLM-generation of formally verified code from formal specifications - in contrast to vibe coding, which generates potentia…