Word Count on LIMA (test)
97.49Surrogate Acc (Initial Constitution)GPT 5-mini
Evaluation Results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| GPT 5-miniTarget Model=GPT 5-mini2026.01 | 97.49 | 97.49 | — | — | 214 | 0.04 | — | — | |
| Gemini 2.5 ProTarget Model=Gemini 2.5 Pro2026.01 | 94.7 | 94.7 | — | — | 193 | 0.04 | — | — | |
| GPT 5Target Model=GPT 52026.01 | 93.85 | 96.08 | — | — | 163 | 0.04 | — | — | |
| GPT 5-nanoTarget Model=GPT 5-nano2026.01 | 93.3 | 95.81 | — | — | 177 | 0.04 | — | — | |
| GPT 4oTarget Model=GPT 4o2026.01 | 93.29 | 95.81 | — | — | 65 | 0.04 | — | — | |
| o4-miniTarget Model=o4-mini2026.01 | 91.92 | 92.47 | — | — | 100 | 0.04 | — | — | |
| Gemini 2.5 FlashTarget Model=Gemini 2.5 Flash2026.01 | 87.46 | 87.46 | — | — | 90 | 0.04 | — | — |