Semantic Textual Similarity on SICK-R (test)
66Similarity ScoreRAIR-Laplacian
Evaluation Results
| Method | Links | |
|---|---|---|
| RAIR-LaplacianHuman comparisons=6002026.06 | 66 | |
| Random EloHuman comparisons=6002026.06 | 63.9 | |
| SGSHuman comparisons=600, Inferred comparisons=5352026.06 | 63.5 | |
| Active EloHuman comparisons=6002026.06 | 62.5 | |
| Active Elo (same budget)Human comparisons=6002026.06 | 62.3 | |
| Cayley-EncoderBase Model=Llama3-8B, GNN Variant=GIN2026.03 | 60.74 | |
| Cayley-EncoderBase Model=Gemma2-2B, GNN Variant=GIN2026.03 | 60.32 | |
| FC-EncoderBase Model=Gemma2-2B, GNN Variant=GCN2026.03 | 60.12 | |
| Best-LayerBase Model=Pythia-410m2026.03 | 58.26 | |
| Cayley-EncoderBase Model=Gemma2-2B, GNN Variant=GCN2026.03 | 57.52 | |
| FC-EncoderBase Model=Llama3-8B, GNN Variant=GCN2026.03 | 57.51 | |
| Best-LayerBase Model=Gemma2-2B2026.03 | 57.24 | |
| FC-EncoderBase Model=Pythia-410m, GNN Variant=GCN2026.03 | 57.14 | |
| Best-LayerBase Model=Llama3-8B2026.03 | 56.96 | |
| WeightedBase Model=Llama3-8B2026.03 | 56.91 | |
| DWAttBase Model=Gemma2-2B2026.03 | 55.36 | |
| Cayley-EncoderBase Model=Pythia-410m, GNN Variant=GIN2026.03 | 55.34 | |
| WeightedBase Model=Pythia-410m2026.03 | 55.09 | |
| FC-EncoderBase Model=Gemma2-2B, GNN Variant=GIN2026.03 | 54.67 | |
| Last-LayerBase Model=Pythia-410m2026.03 | 52.55 | |
| DWAttBase Model=Pythia-410m2026.03 | 52.43 | |
| Cayley-EncoderBase Model=Pythia-410m, GNN Variant=GCN2026.03 | 52.15 | |
| FC-EncoderBase Model=Llama3-8B, GNN Variant=GIN2026.03 | 51.85 | |
| DWAttBase Model=Llama3-8B2026.03 | 51.45 | |
| Last-LayerBase Model=Llama3-8B2026.03 | 51.16 | |
| FC-EncoderBase Model=Pythia-410m, GNN Variant=GIN2026.03 | 46.97 | |
| WeightedBase Model=Gemma2-2B2026.03 | 46.67 | |
| MLP Last-LayerBase Model=Llama3-8B2026.03 | 46.62 | |
| Set-EncoderBase Model=Llama3-8B2026.03 | 46.6 | |
| Set-EncoderBase Model=Pythia-410m2026.03 | 44.84 | |
| Cayley-EncoderBase Model=Llama3-8B, GNN Variant=GCN2026.03 | 44.04 | |
| MLP Last-LayerBase Model=Pythia-410m2026.03 | 43.23 | |
| Last-LayerBase Model=Gemma2-2B2026.03 | 43.01 | |
| Set-EncoderBase Model=Gemma2-2B2026.03 | 42.22 | |
| MLP Last-LayerBase Model=Gemma2-2B2026.03 | 40.19 | |
| PairS-MSHuman comparisons=6002026.06 | 38.2 |