Mathematical Problem Solving on AIME
1,289.8AIME ScoreMSV 64
Evaluation Results
| Method | Links | |
|---|---|---|
| MSV 64Budget (N)=64, Sequence Scale=642026.03 | 1,289.8 | |
| Probe + WV 64Budget (N)=64, Weighted Voting=true2026.03 | 1,167.81 | |
| ProbeBudget (N)=642026.03 | 1,149.4 | |
| MSV 16Budget (N)=16, Sequence Scale=162026.03 | 1,078.29 | |
| MSV 1 + WV 64Budget (N)=64, Sequence Scale=1, Weighted Voting=true2026.03 | 1,045.05 | |
| ProbeBudget (N)=162026.03 | 1,024.18 | |
| MSV 1Budget (N)=64, Sequence Scale=12026.03 | 1,005.34 | |
| Probe + WV 16Budget (N)=16, Weighted Voting=true2026.03 | 1,000.25 | |
| MSV 1 + WV 16Budget (N)=16, Sequence Scale=1, Weighted Voting=true2026.03 | 959.76 | |
| MSV 1Budget (N)=16, Sequence Scale=12026.03 | 954.76 | |
| MSV 4Budget (N)=4, Sequence Scale=42026.03 | 850.44 | |
| ProbeBudget (N)=42026.03 | 823.84 | |
| Probe + WV 4Budget (N)=4, Weighted Voting=true2026.03 | 808.66 | |
| MSV 1Budget (N)=4, Sequence Scale=12026.03 | 793.35 | |
| MSV 1 + WV 4Budget (N)=4, Sequence Scale=1, Weighted Voting=true2026.03 | 789.28 | |
| ProbeBudget (N)=12026.03 | 563.44 | |
| MSV 1Budget (N)=1, Sequence Scale=12026.03 | 551.56 | |
| DeepSeek-R1Model Category=Closed-source Models2026.01 | 86.7 | |
| ReMemFrozen Executor=Qwen3-8B-Thinking, Parameters=8B2026.02 | 61.67 | |
| UMEM-Llama-3.2-1BFrozen Executor=Qwen3-8B-Thinking, Parameters=1B2026.02 | 60 | |
| UMEM-Qwen3-4BFrozen Executor=Gemini-2.5-Flash, Parameters=4B2026.02 | 60 | |
| UMEM-Qwen3-4BFrozen Executor=Qwen3-8B-Thinking, Parameters=4B2026.02 | 58.33 | |
| UMEM-Llama-3.2-1BFrozen Executor=Gemini-2.5-Flash, Parameters=1B2026.02 | 58.33 | |
| o1-miniModel Category=Closed-source Models2026.01 | 56.7 | |
| LoGos(32B)Model Category=Open-source Models (32B), Parameter Count=32B2026.01 | 56.7 | |
| Self-RAGFrozen Executor=Gemini-2.5-Flash, Parameters=API2026.02 | 56.67 | |
| ReMemFrozen Executor=Gemini-2.5-Flash, Parameters=API2026.02 | 56.67 | |
| Qwen3-4B-InstructFrozen Executor=Gemini-2.5-Flash, Parameters=4B2026.02 | 56.67 | |
| Qwen3-4B-InstructFrozen Executor=Qwen3-8B-Thinking, Parameters=4B2026.02 | 55 | |
| No MemoryFrozen Executor=Gemini-2.5-Flash, Parameters=-2026.02 | 53.33 | |
| MempFrozen Executor=Gemini-2.5-Flash, Parameters=API2026.02 | 53.33 | |
| No MemoryFrozen Executor=Qwen3-8B-Thinking, Parameters=-2026.02 | 51.67 | |
| Llama-3.2-1B-InstructFrozen Executor=Qwen3-8B-Thinking, Parameters=1B2026.02 | 51.67 | |
| UMEM-Qwen3-4BFrozen Executor=GPT-5.1, Parameters=4B2026.02 | 51.67 | |
| Llama-3.2-1B-InstructFrozen Executor=Gemini-2.5-Flash, Parameters=1B2026.02 | 51.67 | |
| Self-RAGFrozen Executor=GPT-5.1, Parameters=API2026.02 | 50 | |
| DeepSeek-R1-Distill-Qwen-32BModel Category=Open-source Models (32B), Parameter Count=32B, Model Variant=Distill2026.01 | 46.7 | |
| MempFrozen Executor=Qwen3-8B-Thinking, Parameters=8B2026.02 | 46.67 | |
| Qwen3-4B-InstructFrozen Executor=GPT-5.1, Parameters=4B2026.02 | 46.67 | |
| MempFrozen Executor=GPT-5.1, Parameters=API2026.02 | 45 | |
| UMEM-Llama-3.2-1BFrozen Executor=GPT-5.1, Parameters=1B2026.02 | 45 | |
| Llama-3.2-1B-InstructFrozen Executor=GPT-5.1, Parameters=1B2026.02 | 43.33 | |
| No MemoryFrozen Executor=GPT-5.1, Parameters=-2026.02 | 40 | |
| LoGos(7B)Model Category=Open-source Models (7B), Parameter Count=7B2026.01 | 40 | |
| DeepSeek-R1-Distill-Qwen-7BModel Category=Open-source Models (7B), Parameter Count=7B, Model Variant=Distill2026.01 | 33.3 | |
| Self-RAGFrozen Executor=Qwen3-8B-Thinking, Parameters=8B2026.02 | 30 | |
| ReMemFrozen Executor=GPT-5.1, Parameters=API2026.02 | 30 | |
| Claude3.7-SonnetModel Category=Closed-source Models2026.01 | 30 | |
| Qwen2.5-32B-InstructModel Category=Open-source Models (32B), Parameter Count=32B, Model Variant=Instruct2026.01 | 20 | |
| Qwen2.5-7B-BaseModel Category=Open-source Models (7B), Parameter Count=7B, Model Variant=Base2026.01 | 13.3 | |
| Qwen2.5-32B-BaseModel Category=Open-source Models (32B), Parameter Count=32B, Model Variant=Base2026.01 | 10 | |
| Qwen2.5-7B-InstructModel Category=Open-source Models (7B), Parameter Count=7B, Model Variant=Instruct2026.01 | 3.3 |