Entities
model
GPT-6
- 2026-11-04benchmarkNeutralIMP 70Researchers Introduce SEVERUS Benchmark to Evaluate Coding Agent Proof Generation