Vero shows the gap: agents rarely build reusable lemma libraries (74% of proof lines)

dawnsongtweets · x · 2026-08-23

Vero results reveal where the capability gap lies: the strongest agent passes 87% of individual specifications but leaves 16 repositories unfinished—remaining specs encode cross-module invariants requiring reusable lemma libraries, which agents rarely build. Across 82 full solves, 74% of proof lines live in helper lemmas shared across specifications. Agents also sometimes abandoned hard-to-verify reference algorithms and wrote simpler implementations satisfying the same specs.

Related event: Vero Benchmark: GPT-5.5 Fully Verifies Only 27/43 Repositories(2 posts)→

Original post →

More from Research

Research channel →