Formal Mathematical Answer-Construction on MathArena 2026
6AIME I ScoreGPT-5.4
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| GPT-5.4Pass@k=322025.05 | 6 | 2 | 5 | 0 | 13 | |
| GPT-5.4+Goedel-Prover-V2-8BPass@k=42025.05 | 6 | 4 | 6 | 0 | 16 | |
| GPT-5.4+Goedel-Prover-V2-8BPass@k=322025.05 | 6 | 4 | 7 | 0 | 17 | |
| GPT-5.4+Goedel-Prover-V2-32BPass@k=42025.05 | 6 | 4 | 7 | 0 | 17 | |
| GPT-5.4+Goedel-Prover-V2-32BPass@k=322025.05 | 6 | 5 | 7 | 0 | 18 | |
| Goedel-Prover-V2-32BPass@k=322025.05 | 4 | 1 | 2 | 0 | 7 | |
| Goedel-Prover-V2-32BPass@k=42025.05 | 3 | 1 | 2 | 0 | 6 | |
| GPT-5.4Pass@k=42025.05 | 2 | 2 | 3 | 0 | 7 |