Self-optimization on OPT-BENCH-ML
18Win Countgpt-4o-2024-08-06
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| gpt-4o-2024-08-06Steps=202026.05 | 18 | 1.89 | 61 | |
| grok-3Steps=202026.05 | 15 | 1.29 | 63 | |
| Qwen2.5-72B-InstructSteps=52026.05 | 15 | 1.58 | 40 | |
| Qwen2.5-72B-InstructSteps=102026.05 | 15 | 1.47 | 43 | |
| Qwen2.5-72B-InstructSteps=202026.05 | 15 | 1.61 | 45 | |
| gpt-4.1-2025-04-14Steps=102026.05 | 14 | 1.67 | 50 | |
| gpt-4.1-2025-04-14Steps=202026.05 | 14 | 2.15 | 58 | |
| gpt-o3-miniSteps=102026.05 | 14 | 1.77 | 63 | |
| gemini-2.0-flashSteps=102026.05 | 14 | 2.08 | 48 | |
| gemini-2.0-flashSteps=202026.05 | 14 | 1.95 | 53 | |
| claude-3-7-sonnet-20250219Steps=52026.05 | 14 | 1 | 43 | |
| claude-3-7-sonnet-20250219Steps=102026.05 | 14 | 1.11 | 48 | |
| claude-3-7-sonnet-20250219Steps=202026.05 | 14 | 1.35 | 53 | |
| grok-3Steps=102026.05 | 14 | 1.36 | 55 | |
| Qwen3-235B-ThinkingSteps=202026.05 | 14 | 1.13 | 62 | |
| gpt-4o-2024-08-06Steps=52026.05 | 13 | 1.28 | 53 | |
| gpt-4o-2024-08-06Steps=102026.05 | 13 | 1.8 | 58 | |
| gpt-o3-miniSteps=202026.05 | 13 | 1.35 | 65 | |
| gemini-2.0-flashSteps=52026.05 | 13 | 1.08 | 45 | |
| Qwen3-235B-ThinkingSteps=102026.05 | 13 | 1.13 | 55 | |
| Qwen3-235B-InstructSteps=102026.05 | 13 | 1.18 | 54 | |
| Qwen3-235B-InstructSteps=202026.05 | 13 | 1.03 | 60 | |
| gpt-4.1-2025-04-14Steps=52026.05 | 12 | 1.11 | 40 | |
| gpt-o3-miniSteps=52026.05 | 12 | 1.25 | 55 | |
| claude-3-5-sonnet-20241022Steps=102026.05 | 12 | 1.45 | 40 | |
| claude-3-5-sonnet-20241022Steps=202026.05 | 12 | 1.83 | 48 | |
| grok-3Steps=52026.05 | 12 | 1.16 | 50 | |
| Deepseek-V3.1-ThinkingSteps=52026.05 | 12 | 1.13 | 50 | |
| Deepseek-V3.1-ThinkingSteps=102026.05 | 12 | 1.18 | 58 | |
| Qwen3-235B-ThinkingSteps=52026.05 | 12 | 1.18 | 52 | |
| Qwen3-235B-InstructSteps=52026.05 | 12 | 1.13 | 48 | |
| Qwen2.5-14B-InstructSteps=202026.05 | 12 | 1.13 | 30 | |
| Qwen3-8BSteps=202026.05 | 12 | 1.16 | 38 | |
| Qwen3-30B-A3BSteps=202026.05 | 12 | 1.14 | 54 | |
| Qwen3-32BSteps=202026.05 | 12 | 1.28 | 58 | |
| claude-3-5-sonnet-20241022Steps=52026.05 | 11 | 1.15 | 30 | |
| Deepseek-V3.1-ThinkingSteps=202026.05 | 11 | 1.03 | 60 | |
| Qwen2.5-14B-InstructSteps=102026.05 | 11 | 1.11 | 25 | |
| Qwen2.5-32B-InstructSteps=202026.05 | 11 | 1.19 | 45 | |
| Qwen3-30B-A3BSteps=102026.05 | 11 | 1.08 | 53 | |
| Qwen3-32BSteps=102026.05 | 11 | 1.19 | 48 | |
| Internlm3-8b-instructSteps=202026.05 | 10 | 1.01 | 30 | |
| Qwen2.5-14B-InstructSteps=52026.05 | 10 | 1.39 | 18 | |
| Qwen2.5-32B-InstructSteps=102026.05 | 10 | 1.21 | 33 | |
| Qwen3-8BSteps=102026.05 | 10 | 1.24 | 33 | |
| Qwen3-30B-A3BSteps=52026.05 | 10 | 1.06 | 45 | |
| Qwen3-32BSteps=52026.05 | 10 | 1.07 | 18 | |
| Internlm3-8b-instructSteps=102026.05 | 9 | 0.98 | 23 | |
| Qwen2.5-7B-InstructSteps=52026.05 | 9 | 0.99 | 3 | |
| Qwen2.5-32B-InstructSteps=52026.05 | 9 | 1.38 | 25 | |
| Qwen3-8BSteps=52026.05 | 9 | 0.99 | 28 | |
| Internlm3-8b-instructSteps=52026.05 | 8 | 0.94 | 5 | |
| Qwen2.5-3B-InstructSteps=52026.05 | 8 | 0.98 | 5 | |
| Qwen2.5-3B-InstructSteps=102026.05 | 7 | 0.85 | 13 | |
| Qwen2.5-7B-InstructSteps=102026.05 | 7 | 0.82 | 15 | |
| Qwen2.5-3B-InstructSteps=202026.05 | 6 | 0.63 | 15 | |
| Qwen2.5-7B-InstructSteps=202026.05 | 6 | 0.63 | 20 |