Benchmark RadarAI BENCHMARK PROFILE
AdvancedMathBench
AdvancedMathBench is a benchmark suite for advanced mathematical reasoning, containing ProverBench (296 proof problems) and VerifierBench (888 proof trajectories with expert labels), with an automatic verification pipeline.
- Released
- 2026-07-13
- Readiness
- Paper only
- Primary field
- General AI
Why it matters
Existing math benchmarks focus on high-school levels and final answers. AdvancedMathBench evaluates proof generation and verification at advanced levels, providing fine-grained assessments of proof correctness and error detection.
Motivation
Large language models (LLMs) have achieved remarkable performance on high-school and olympiad-style mathematics, yet their capabilities on advanced mathematics remain poorly understood.
Primary resources
Benchmark Radar records only publicly supported details and links back to primary sources for verification.