| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| AdvancedMath | Kimina-7B | Compilation Pass Rate@10100 | 28 | 4mo ago | |
| ProofNet | Kimina-7B | Compilation Pass Rate@1094.1 | 28 | 4mo ago | |
| MiniF2F | Godel-V2-8B | Compilation Pass Rate@10100 | 28 | 4mo ago | |
| ConNF | DRIFT | TC@172.32 | 16 | 3mo ago | |
| miniF2F (test) | TC@196 | 16 | 3mo ago | ||
| ProofNet (test) | Monotonic Reference-Free Refinement | πFV44.09 | 12 | 3mo ago | |
| Gaokao Formal | SFT+GRPO-0% | Mean Score74.2 | 8 | 3mo ago | |
| PRIME | DSR | Success Count (SC)80.13 | 7 | 3mo ago | |
| ProverBench | Goedel-V2-Formalizer-32B | Success Count95.38 | 7 | 3mo ago | |
| PutnamBench (PB) | RL (GRPO) 2B | Mean Cycle Consistency0.561 | 6 | 4mo ago | |
| FLC (held-out) | RL (GRPO) 2B | Mean Cycle Consistency66.9 | 6 | 4mo ago | |
| FLC (val) | SFT No-Curriculum 2B | Cross-Entropy Loss0.64 | 3 | 4mo ago | |
| ESS #1196 | LeanMarathon | Proof Success Rate100 | 2 | 1mo ago | |
| Erdős–Graham | LeanMarathon | Targets Proven3 | 2 | 1mo ago | |
| Munkres’ Topology (Sections 12–50 (39)) | Isabelle/HOL | Active days24 | 2 | 3mo ago |