Scientific Knowledge Evaluation on SciKnowEval
52.1Accuracyo3
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| o3Reasoning Effort=Low2025.08 | 52.1 | — | |
| o3Reasoning Effort=High2025.08 | 51.9 | -0.2 | |
| o3-miniReasoning Effort=High2025.08 | 51.9 | 2.9 | |
| o4-miniReasoning Effort=High2025.08 | 51.1 | 1.2 | |
| o4-miniReasoning Effort=Low2025.08 | 49.9 | — | |
| o3-miniReasoning Effort=Low2025.08 | 49 | — | |
| Gemini-2.5-ProReasoning Effort=High2025.08 | 47.6 | 0.8 | |
| Gemini-2.5-ProReasoning Effort=Low2025.08 | 46.8 | — | |
| GPT-5Reasoning Effort=High2025.08 | 46.7 | 1.2 | |
| GPT-5Reasoning Effort=Low2025.08 | 45.5 | — | |
| Claude-Sonnet-4Reasoning Effort=Low2025.08 | 43.6 | — | |
| Claude-Sonnet-4Reasoning Effort=High2025.08 | 43.3 | -0.3 |