Factual Consistency Evaluation on QAGS Average (test)
0.604Pearson rBINEVAL
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| BINEVALBackbone=Claude2026.06 | 0.604 | 0.62 | 0.534 | |
| G-EvalBackbone=GPT-42026.06 | 0.599 | 0.611 | 0.525 | |
| UniEvalBackbone=T52026.06 | 0.571 | 0.575 | 0.465 | |
| BINEVALBackbone=gpt-oss2026.06 | 0.543 | 0.563 | 0.492 | |
| CTC2026.06 | 0.464 | 0.43 | 0.346 | |
| BARTScore2026.06 | 0.459 | 0.42 | 0.343 | |
| UniEvalBackbone=gpt-oss2026.06 | 0.452 | 0.436 | 0.424 | |
| QAGS2026.06 | 0.375 | — | — | |
| FactCC2026.06 | 0.356 | 0.371 | 0.294 | |
| BERTScore2026.06 | 0.3 | 0.256 | 0.202 | |
| MoverScore2026.06 | 0.234 | 0.195 | 0.153 | |
| ROUGE-L2026.06 | 0.19 | 0.156 | 0.122 | |
| G-EvalBackbone=gpt-oss2026.06 | 0.14 | 0.132 | 0.131 |