Evidence Retrieval on Factiverse Evaluation Dataset
82.16F1 ScoreXLM-RoBERTa-Large (FT)
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| XLM-RoBERTa-Large (FT)Provider=Factiverse2026.06 | 82.16 | 94.29 | 72.79 | |
| pplx-embed-v1-4bProvider=Perplexity2026.06 | 81.77 | 79.05 | 84.69 | |
| pplx-embed-v1-0.6bProvider=Perplexity2026.06 | 79.79 | 71.43 | 90.36 | |
| text-embedding-3-largeProvider=OpenAI2026.06 | 79.59 | 74.29 | 85.71 | |
| Qwen3-Embedding-0.6BProvider=Qwen2026.06 | 76.52 | 96.19 | 63.52 | |
| Qwen3-Embedding-4BProvider=Qwen2026.06 | 75 | 88.57 | 65.03 | |
| embeddinggemma-300mProvider=Google2026.06 | 53.41 | 89.52 | 38.06 |