Latest updates from Boundless Intuition Labs.
Benchmarks, verification results, and the failures we found along the way, published as we finish a report, not on a schedule.
Dirac Achieves a Perfect 6/6 on IMO 2026 at Record Speed
We ran Dirac, our autonomous proving agent, on the publicly released formalizations of all six IMO 2026 problems. It produced a machine-checked proof of every one, in 7h 18m of total proving time at a cost of $176.58.
Towards Verified Superintelligence
AI is becoming the operating system of the modern world, yet we have no reliable way to know whether an AI system is actually correct. Scaling intelligence without scaling trust is a dangerous trajectory.
Fluency Is Not Correctness
On RuleArena's airline domain, verification lifts two frontier Claude models from 54% and 61% to 100% while cutting cost roughly fourteenfold, and a verified budget model beats both unaided frontier models.