Reasoning on LiveBench Reasoning
92AccuracyDIP
Evaluation results
| Method | Links | ||
|---|---|---|---|
| DIPBackbone=Grok 3 Mini (High), Prompting Strategy=Diverge-to-Induce Prompting2026.02 | 92 | 7.5 | |
| DIPBackbone=o4 Mini, Prompting Strategy=Diverge-to-Induce Prompting2026.02 | 91 | 8.5 | |
| S-CoTBackbone=Grok 3 Mini (High), Prompting Strategy=Step-by-step CoT2026.02 | 86.5 | — | |
| Z-CoTBackbone=Grok 3 Mini (High), Prompting Strategy=Zero-shot CoT2026.02 | 84.5 | — | |
| DIPBackbone=Grok 3 Mini (Low), Prompting Strategy=Diverge-to-Induce Prompting2026.02 | 84.5 | 2.5 | |
| S-CoTBackbone=o4 Mini, Prompting Strategy=Step-by-step CoT2026.02 | 84 | — | |
| S-CoTBackbone=Grok 3 Mini (Low), Prompting Strategy=Step-by-step CoT2026.02 | 84 | — | |
| R-CoTBackbone=Grok 3 Mini (High), Prompting Strategy=Rationalized CoT2026.02 | 83 | — | |
| Z-CoTBackbone=o4 Mini, Prompting Strategy=Zero-shot CoT2026.02 | 82.5 | — | |
| Z-CoTBackbone=Grok 3 Mini (Low), Prompting Strategy=Zero-shot CoT2026.02 | 82 | — | |
| R-CoTBackbone=Grok 3 Mini (Low), Prompting Strategy=Rationalized CoT2026.02 | 81 | — | |
| R-CoTBackbone=o4 Mini, Prompting Strategy=Rationalized CoT2026.02 | 80 | — | |
| S-CoTBackbone=Grok 3, Prompting Strategy=Step-by-step CoT2026.02 | 77 | — | |
| Z-CoTBackbone=Grok 3, Prompting Strategy=Zero-shot CoT2026.02 | 75 | — | |
| DIPBackbone=Grok 3, Prompting Strategy=Diverge-to-Induce Prompting2026.02 | 73 | -2 | |
| R-CoTBackbone=Grok 3, Prompting Strategy=Rationalized CoT2026.02 | 72 | — | |
| DIPBackbone=GPT 4.1, Prompting Strategy=Diverge-to-Induce Prompting2026.02 | 70.5 | 5 | |
| Z-CoTBackbone=GPT 4.1, Prompting Strategy=Zero-shot CoT2026.02 | 65.5 | — | |
| DIPBackbone=GPT 4.1 Mini, Prompting Strategy=Diverge-to-Induce Prompting2026.02 | 65 | 13 | |
| R-CoTBackbone=GPT 4.1, Prompting Strategy=Rationalized CoT2026.02 | 64.5 | — | |
| S-CoTBackbone=GPT 4.1 Mini, Prompting Strategy=Step-by-step CoT2026.02 | 63 | — | |
| S-CoTBackbone=GPT 4.1, Prompting Strategy=Step-by-step CoT2026.02 | 63 | — | |
| DIPBackbone=Gemini 2.0 Flash, Prompting Strategy=Diverge-to-Induce Prompting2026.02 | 58.5 | 4 | |
| DIPBackbone=Llama 4 Maverick, Prompting Strategy=Diverge-to-Induce Prompting2026.02 | 56.5 | 1.5 | |
| Z-CoTBackbone=Llama 4 Maverick, Prompting Strategy=Zero-shot CoT2026.02 | 55 | — | |
| S-CoTBackbone=Gemini 2.0 Flash, Prompting Strategy=Step-by-step CoT2026.02 | 55 | — | |
| DIPBackbone=Mistral Medium 3, Prompting Strategy=Diverge-to-Induce Prompting2026.02 | 54.5 | 1.5 | |
| Z-CoTBackbone=Gemini 2.0 Flash, Prompting Strategy=Zero-shot CoT2026.02 | 54.5 | — | |
| R-CoTBackbone=GPT 4.1 Mini, Prompting Strategy=Rationalized CoT2026.02 | 54.5 | — | |
| S-CoTBackbone=Mistral Medium 3, Prompting Strategy=Step-by-step CoT2026.02 | 53.5 | — | |
| Z-CoTBackbone=Mistral Medium 3, Prompting Strategy=Zero-shot CoT2026.02 | 53 | — | |
| R-CoTBackbone=Gemini 2.0 Flash, Prompting Strategy=Rationalized CoT2026.02 | 52 | — | |
| Z-CoTBackbone=GPT 4.1 Mini, Prompting Strategy=Zero-shot CoT2026.02 | 52 | — | |
| S-CoTBackbone=Llama 4 Maverick, Prompting Strategy=Step-by-step CoT2026.02 | 50 | — | |
| R-CoTBackbone=Llama 4 Maverick, Prompting Strategy=Rationalized CoT2026.02 | 49.5 | — | |
| R-CoTBackbone=Mistral Medium 3, Prompting Strategy=Rationalized CoT2026.02 | 49 | — | |
| DIPBackbone=Gemini 2.0 Flash Lite, Prompting Strategy=Diverge-to-Induce Prompting2026.02 | 48 | 9.5 | |
| DIPBackbone=Llama 3.3 70B, Prompting Strategy=Diverge-to-Induce Prompting2026.02 | 47 | 5 | |
| DIPBackbone=Mistral Large 2, Prompting Strategy=Diverge-to-Induce Prompting2026.02 | 46.5 | 2 | |
| R-CoTBackbone=GPT 4o, Prompting Strategy=Rationalized CoT2026.02 | 46.5 | — | |
| DIPBackbone=GPT 4o, Prompting Strategy=Diverge-to-Induce Prompting2026.02 | 46 | 4.5 | |
| Z-CoTBackbone=Mistral Large 2, Prompting Strategy=Zero-shot CoT2026.02 | 44.5 | — | |
| S-CoTBackbone=Llama 3.3 70B, Prompting Strategy=Step-by-step CoT2026.02 | 44 | — | |
| S-CoTBackbone=Mistral Large 2, Prompting Strategy=Step-by-step CoT2026.02 | 43.5 | — | |
| DIPBackbone=Llama 4 Scout, Prompting Strategy=Diverge-to-Induce Prompting2026.02 | 43 | 30.5 | |
| R-CoTBackbone=Gemini 2.0 Flash Lite, Prompting Strategy=Rationalized CoT2026.02 | 43 | — | |
| Z-CoTBackbone=Llama 3.3 70B, Prompting Strategy=Zero-shot CoT2026.02 | 42 | — | |
| R-CoTBackbone=Mistral Large 2, Prompting Strategy=Rationalized CoT2026.02 | 42 | — | |
| S-CoTBackbone=GPT 4o, Prompting Strategy=Step-by-step CoT2026.02 | 42 | — | |
| Z-CoTBackbone=GPT 4o, Prompting Strategy=Zero-shot CoT2026.02 | 41.5 | — | |
| DIPBackbone=Mistral Small 3.1, Prompting Strategy=Diverge-to-Induce Prompting2026.02 | 39.5 | 0.5 | |
| Z-CoTBackbone=Mistral Small 3.1, Prompting Strategy=Zero-shot CoT2026.02 | 39 | — | |
| S-CoTBackbone=Gemini 2.0 Flash Lite, Prompting Strategy=Step-by-step CoT2026.02 | 39 | — | |
| Z-CoTBackbone=Gemini 2.0 Flash Lite, Prompting Strategy=Zero-shot CoT2026.02 | 38.5 | — | |
| S-CoTBackbone=Mistral Small 3.1, Prompting Strategy=Step-by-step CoT2026.02 | 38 | — | |
| Z-CoTBackbone=Mistral Small 3, Prompting Strategy=Zero-shot CoT2026.02 | 36.5 | — | |
| DIPBackbone=Mistral Small 3, Prompting Strategy=Diverge-to-Induce Prompting2026.02 | 36.5 | 0 | |
| R-CoTBackbone=Mistral Small 3, Prompting Strategy=Rationalized CoT2026.02 | 36 | — | |
| S-CoTBackbone=Mistral Small 3, Prompting Strategy=Step-by-step CoT2026.02 | 35.5 | — | |
| Z-CoTBackbone=GPT 4.1 Nano, Prompting Strategy=Zero-shot CoT2026.02 | 35 | — | |
| R-CoTBackbone=Mistral Small 3.1, Prompting Strategy=Rationalized CoT2026.02 | 33.5 | — | |
| DIPBackbone=GPT 4o mini, Prompting Strategy=Diverge-to-Induce Prompting2026.02 | 32.5 | 3 | |
| DIPBackbone=GPT 4.1 Nano, Prompting Strategy=Diverge-to-Induce Prompting2026.02 | 32 | -3 | |
| DIPBackbone=Mistral Small 2, Prompting Strategy=Diverge-to-Induce Prompting2026.02 | 31 | 1.5 | |
| S-CoTBackbone=GPT 4o mini, Prompting Strategy=Step-by-step CoT2026.02 | 31 | — | |
| R-CoTBackbone=Mistral Small 2, Prompting Strategy=Rationalized CoT2026.02 | 30.5 | — | |
| Z-CoTBackbone=Mistral Small 2, Prompting Strategy=Zero-shot CoT2026.02 | 29.5 | — | |
| Z-CoTBackbone=GPT 4o mini, Prompting Strategy=Zero-shot CoT2026.02 | 29.5 | — | |
| R-CoTBackbone=GPT 4.1 Nano, Prompting Strategy=Rationalized CoT2026.02 | 29 | — | |
| S-CoTBackbone=GPT 4.1 Nano, Prompting Strategy=Step-by-step CoT2026.02 | 29 | — | |
| DIPBackbone=Llama 3.3 8B, Prompting Strategy=Diverge-to-Induce Prompting2026.02 | 28 | 6.5 | |
| S-CoTBackbone=Mistral Small 2, Prompting Strategy=Step-by-step CoT2026.02 | 28 | — | |
| R-CoTBackbone=GPT 4o mini, Prompting Strategy=Rationalized CoT2026.02 | 24.5 | — | |
| R-CoTBackbone=Llama 3.3 70B, Prompting Strategy=Rationalized CoT2026.02 | 22.5 | — | |
| Z-CoTBackbone=Llama 3.3 8B, Prompting Strategy=Zero-shot CoT2026.02 | 21.5 | — | |
| S-CoTBackbone=Llama 3.3 8B, Prompting Strategy=Step-by-step CoT2026.02 | 18.5 | — | |
| R-CoTBackbone=Llama 3.3 8B, Prompting Strategy=Rationalized CoT2026.02 | 16.5 | — | |
| S-CoTBackbone=Llama 4 Scout, Prompting Strategy=Step-by-step CoT2026.02 | 13.5 | — | |
| Z-CoTBackbone=Llama 4 Scout, Prompting Strategy=Zero-shot CoT2026.02 | 12.5 | — | |
| R-CoTBackbone=Llama 4 Scout, Prompting Strategy=Rationalized CoT2026.02 | 8 | — |