Mathematical Reasoning on Beyond AIME (Performance %, Tokens)
1Total TokensTopoDIM
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| TopoDIMBackbone=GPT-OSS:120B2026.04 | 1 | — | |
| GTDBackbone=GPT-OSS:120B2026.04 | 1.17 | — | |
| G-designerBackbone=GPT-OSS:120B2026.04 | 1.21 | — | |
| LongGraphBackbone=GPT-OSS:120B2026.04 | 1.29 | — | |
| AutoGenBackbone=GPT-OSS:120B2026.04 | 1.85 | — | |
| AgentFmwBackbone=GPT-OSS:120B2026.04 | 1.85 | — | |
| GPT-SwarmBackbone=GPT-OSS:120B2026.04 | 2.49 | — | |
| LLM-DEBATEBackbone=GPT-OSS:120B2026.04 | 2.68 | — | |
| LobsterBackbone=GPT-OSS:120B2026.04 | 108 | — | |
| Agent Q-MixBackbone=GPT-OSS:120B2026.04 | 708 | — |