Data as of Jul 25, 2026 · Based on 2,718,867 AI responses across 9,511 prompts · See how Parse measures this
Paragonpolarity.cc · Low evidencePolarity provides sandboxed eval infrastructure for AI agents, running each task inside an isolated Docker environment with real backing services like Postgres, Redis, and S3. It scores agent runs against behavioral invariants and forbidden rules, measures non-determinism via replicas, and enables one-command local reproduction of failures.
PolarityParagonParse Score