Natural Language Inference on WANLI
45.4Accuracy (WANLI)DOPA*
Evaluation Results
| Method | Links | |
|---|---|---|
| DOPA*Backbone=Qwen3-8B2026.04 | 45.4 | |
| DrICLBackbone=Qwen3-1.7B2026.04 | 42.77 | |
| DICLBackbone=Qwen3-1.7B2026.04 | 42.7 | |
| DOPA*Backbone=Qwen3-1.7B2026.04 | 42.37 | |
| KNNBackbone=Qwen3-8B2026.04 | 41.93 | |
| KNNBackbone=Qwen3-1.7B2026.04 | 41.77 | |
| DICLBackbone=Qwen3-8B2026.04 | 41.6 | |
| DOPA*Backbone=LLaMA3.1-8B2026.04 | 41.33 | |
| RandomBackbone=Qwen3-1.7B2026.04 | 41.3 | |
| RandomBackbone=Qwen3-8B2026.04 | 41.13 | |
| KNNBackbone=LLaMA3.1-8B2026.04 | 41 | |
| InfICLBackbone=LLaMA3.1-8B2026.04 | 40.87 | |
| InfICLBackbone=Qwen3-1.7B2026.04 | 40.8 | |
| DrICLBackbone=LLaMA3.1-8B2026.04 | 40.7 | |
| DrICLBackbone=Qwen3-8B2026.04 | 40.57 | |
| InfICLBackbone=Qwen3-8B2026.04 | 40.07 | |
| RandomBackbone=LLaMA3.1-8B2026.04 | 39.73 | |
| RewriteBackbone=Qwen3-1.7B2026.04 | 39.47 | |
| DICLBackbone=LLaMA3.1-8B2026.04 | 38.97 | |
| RewriteBackbone=LLaMA3.1-8B2026.04 | 38.87 | |
| RewriteBackbone=Qwen3-8B2026.04 | 38.83 | |
| DOPA*Backbone=LLaMA3.2-3B2026.04 | 38.4 | |
| RandomBackbone=LLaMA3.2-3B2026.04 | 37.7 | |
| DICLBackbone=LLaMA3.2-3B2026.04 | 37.67 | |
| KNNBackbone=LLaMA3.2-3B2026.04 | 37.2 | |
| DrICLBackbone=LLaMA3.2-3B2026.04 | 36.5 | |
| RewriteBackbone=LLaMA3.2-3B2026.04 | 36.43 | |
| InfICLBackbone=LLaMA3.2-3B2026.04 | 36.23 | |
| DOPA*Backbone=GPT2-xl2026.04 | 34.93 | |
| RandomBackbone=GPT2-xl2026.04 | 34.23 | |
| RewriteBackbone=GPT2-xl2026.04 | 34 | |
| InfICLBackbone=GPT2-xl2026.04 | 33.8 | |
| DrICLBackbone=Gemma2-2B2026.04 | 33.73 | |
| RewriteBackbone=Gemma2-2B2026.04 | 33.7 | |
| KNNBackbone=GPT2-xl2026.04 | 33.57 | |
| KNNBackbone=Gemma2-2B2026.04 | 33.5 | |
| DICLBackbone=Gemma2-2B2026.04 | 33.43 | |
| RandomBackbone=Gemma2-2B2026.04 | 33.37 | |
| DOPA*Backbone=Gemma2-2B2026.04 | 33.37 | |
| InfICLBackbone=Gemma2-2B2026.04 | 33.27 | |
| DICLBackbone=GPT2-xl2026.04 | 32.93 | |
| DrICLBackbone=GPT2-xl2026.04 | 32.03 |