Last released Jun 19, 2026
Statistically rigorous, batch-first reliability auditing for LLM judges and reward models
Supported by