lf-lean is Theorem’s benchmark showing that frontier AI plus formal methods can translate all 1,276 Logical Foundations statements from Rocq to Lean and formally verify each translation, achieving an estimated 350× speed-up over manual proof development.
Endorsements support Theorem Labs.
lf-lean is Theorem’s benchmark showing that frontier AI plus formal methods can translate all 1,276 Logical Foundations statements from Rocq to Lean and formally verify each translation, achieving an estimated 350× speed-up over manual proof development.
Endorsements support Theorem Labs.
People– no linked people
Updated 05/18/26 · By grantmaking.aiProject Details
Updated 05/18/26 · By grantmaking.ailf-lean is a research benchmark by Theorem in which frontier AI models and Theorem’s verification stack translate all 1,276 theorems and definitions from the Logical Foundations volume of Software Foundations from Rocq to Lean, then mechanically prove each Lean translation equivalent to the original, reducing an estimated 2.7–2.75 person‑years of verification effort to about two person‑days and illustrating roughly 350× speed-ups in verified software engineering workflows.
Grants Received– no grants recorded
Updated 05/18/26 · By grantmaking.aiDiscussion
No comments yet. Be the first to share your thoughts.