Bias Detection on SBIC (test)
93.6AccuracyDeepSeek-R1-32B
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| DeepSeek-R1-32BModel Category=Rule-based LLMs as Judges2025.04 | 93.6 | 0 | |
| Llama-3-8B-itModel Category=Rule-based LLMs as Judges2025.04 | 88 | 0.4 | |
| GPT-4oModel Category=Rule-based LLMs as Judges2025.04 | 80.8 | 5.6 | |
| BiasGuard2025.04 | 74 | 13.2 | |
| Moderation APIModel Category=Bias Classifier2025.04 | 60.4 | 11.6 | |
| Llama-Guard-3Model Category=Bias Classifier2025.04 | 58.4 | 22 | |
| GPT-4oModel Category=LLMs as Judges2025.04 | 58 | 40.4 | |
| Llama-3-8B-itModel Category=LLMs as Judges2025.04 | 56.8 | 42 | |
| ToxigenModel Category=Bias Classifier2025.04 | 55.6 | 38.4 | |
| ShieldGemmaModel Category=Bias Classifier2025.04 | 52 | 9.6 | |
| DeepSeek-R1-32BModel Category=LLMs as Judges2025.04 | 51.6 | 46 | |
| Azure APIModel Category=Bias Classifier2025.04 | 34.69 | 61.22 |