Natural Language Inference on CNLI
47.2AccuracyDrICL
Evaluation Results
| Method | Links | |
|---|---|---|
| DrICLBackbone=GPT2-xl2026.04 | 47.2 | |
| DOPA*Backbone=Gemma2-2B2026.04 | 46.1 | |
| DOPA*Backbone=GPT2-xl2026.04 | 45.77 | |
| DrICLBackbone=Gemma2-2B2026.04 | 45.43 | |
| RewriteBackbone=GPT2-xl2026.04 | 45.38 | |
| RewriteBackbone=Gemma2-2B2026.04 | 45.29 | |
| InfICLBackbone=Gemma2-2B2026.04 | 45.29 | |
| DICLBackbone=Gemma2-2B2026.04 | 45.15 | |
| KNNBackbone=GPT2-xl2026.04 | 45.05 | |
| DICLBackbone=GPT2-xl2026.04 | 44.57 | |
| KNNBackbone=LLaMA3.2-3B2026.04 | 44.24 | |
| RewriteBackbone=LLaMA3.1-8B2026.04 | 43.23 | |
| DOPA*Backbone=LLaMA3.2-3B2026.04 | 43.19 | |
| RandomBackbone=LLaMA3.1-8B2026.04 | 43.09 | |
| InfICLBackbone=LLaMA3.1-8B2026.04 | 43.04 | |
| DrICLBackbone=LLaMA3.2-3B2026.04 | 42.95 | |
| RandomBackbone=LLaMA3.2-3B2026.04 | 42.75 | |
| RewriteBackbone=LLaMA3.2-3B2026.04 | 42.75 | |
| DOPA*Backbone=LLaMA3.1-8B2026.04 | 42.75 | |
| RandomBackbone=Gemma2-2B2026.04 | 42.66 | |
| DICLBackbone=LLaMA3.2-3B2026.04 | 42.52 | |
| DrICLBackbone=LLaMA3.1-8B2026.04 | 42.32 | |
| KNNBackbone=LLaMA3.1-8B2026.04 | 42.23 | |
| DICLBackbone=LLaMA3.1-8B2026.04 | 42.13 | |
| KNNBackbone=Gemma2-2B2026.04 | 41.61 | |
| DOPA*Backbone=Qwen3-1.7B2026.04 | 40.94 | |
| InfICLBackbone=LLaMA3.2-3B2026.04 | 40.65 | |
| InfICLBackbone=Qwen3-1.7B2026.04 | 40.03 | |
| RandomBackbone=GPT2-xl2026.04 | 39.22 | |
| DICLBackbone=Qwen3-1.7B2026.04 | 39.22 | |
| RandomBackbone=Qwen3-1.7B2026.04 | 39.12 | |
| RewriteBackbone=Qwen3-1.7B2026.04 | 38.79 | |
| KNNBackbone=Qwen3-1.7B2026.04 | 37.83 | |
| DrICLBackbone=Qwen3-1.7B2026.04 | 37.49 | |
| InfICLBackbone=Qwen3-8B2026.04 | 35.29 | |
| DOPA*Backbone=Qwen3-8B2026.04 | 34.29 | |
| DrICLBackbone=Qwen3-8B2026.04 | 32.81 | |
| RewriteBackbone=Qwen3-8B2026.04 | 31.09 | |
| KNNBackbone=Qwen3-8B2026.04 | 30.27 | |
| DICLBackbone=Qwen3-8B2026.04 | 29.6 | |
| RandomBackbone=Qwen3-8B2026.04 | 28.17 | |
| InfICLBackbone=GPT2-xl2026.04 | 24.25 |