Autoresearch for SAT Solvers: AI Agents Autonomously Run Formal Methods Experiments — 129pts

GitHub·medium signal

A Show HN post demonstrating AI autoresearch applied to SAT solver development earned 129pts and 24 comments, extending Karpathy's autoresearch pattern from ML into formal verification. The agent autonomously runs solver experiments, interprets results, and iterates — no human in the loop. This signals the autoresearch trend (previously observed in ML benchmarking) is now entering formal methods and combinatorics, where experimental iteration has traditionally been extremely labor-intensive.