Relational Similarity on Relational Similarity Human Judgments Exp. 2
0.831Pearson Correlation (r)Fine-tuned FV
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Fine-tuned FVModel=GPT-J, Intervention Protocol=vz Fine-tuned FV2026.01 | 0.831 | 0.81 | |
| Fine-tuned FVModel=GPT-2, Intervention Protocol=vz Fine-tuned FV2026.01 | 0.822 | 0.8 | |
| Fine-tuned FVModel=LLaMa-3.1, Intervention Protocol=vz Fine-tuned FV2026.01 | 0.749 | 0.72 | |
| Fine-tuned FVModel=LLaMa-2, Intervention Protocol=vz Fine-tuned FV2026.01 | 0.747 | 0.71 | |
| Initial FVModel=GPT-2, Intervention Protocol=vt Initial FV2026.01 | 0.73 | 0.69 | |
| LLaMa-2Intervention Protocol=Baseline2026.01 | 0.73 | 0.69 | |
| Initial FVModel=GPT-J, Intervention Protocol=vt Initial FV2026.01 | 0.668 | 0.63 | |
| Initial FVModel=LLaMa-2, Intervention Protocol=vt Initial FV2026.01 | 0.658 | 0.62 | |
| Initial FVModel=LLaMa-3.1, Intervention Protocol=vt Initial FV2026.01 | 0.645 | 0.6 | |
| LLaMa-3.1Intervention Protocol=Baseline2026.01 | 0.62 | 0.57 | |
| GPT-JIntervention Protocol=Baseline2026.01 | 0.611 | 0.56 | |
| GPT-2Intervention Protocol=Baseline2026.01 | 0.584 | 0.53 |