Operational safety evaluation on OFFTOPICEVAL Hindi 1.0 (test)
98.86AR_IDGPT-OSS (120B)
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| GPT-OSS (120B)Family=GPT2025.09 | 98.86 | 83.33 | 56.22 | 81.81 | |
| Llama-3.3 (70B)Family=Llama-32025.09 | 98.57 | 54.51 | 20.31 | 54.24 | |
| Llama-3.1 (8B)Family=Llama-32025.09 | 98.19 | 19.24 | 2.62 | 19.67 | |
| GPT-OSS (20B)Family=GPT2025.09 | 97.68 | 54.51 | 38.01 | 62.79 | |
| Qwen-3 (235B)Family=Qwen2025.09 | 97.62 | 99.17 | 61.57 | 88.16 | |
| Phi-4 (15B)Family=Phi2025.09 | 92.86 | 77.41 | 49.27 | 59.1 | |
| Gemma-3 (12B)Family=Gemma2025.09 | 83.52 | 69.99 | 10.81 | 18.88 | |
| Qwen-3 (30B)Family=Qwen2025.09 | 71.43 | 93.04 | 52.46 | 72.08 | |
| Gemma-3 (27B)Family=Gemma2025.09 | 71.14 | 63.06 | 18.3 | 29.13 | |
| Ministral (8B)Family=Mistral2025.09 | 59.52 | 66.09 | 25.72 | 51.83 | |
| Mistral (24B)Family=Mistral2025.09 | 48.95 | 99.72 | 96.8 | 65.37 | |
| Phi-4-mini (4B)Family=Phi2025.09 | 48.1 | 54.9 | 98.36 | 64.04 |