Natural Language Inference on Cardiff Biology
35.05NLI AccuracyRLearner-LLM
Evaluation Results
| Method | Links | |
|---|---|---|
| RLearner-LLMBase Model=Gemma 4-E4B-it, Training Phase=RLearner2026.05 | 35.05 | |
| RLearner-LLMBase Model=LLaMA-2-13B, Training Phase=RLearner2026.05 | 32.09 | |
| Gemma 4-E4B-itTraining Phase=SFT2026.05 | 21.17 | |
| Qwen3-8BTraining Phase=SFT2026.05 | 19.59 | |
| RLearner-LLMBase Model=Qwen3-8B, Training Phase=RLearner2026.05 | 18.2 | |
| ILearner-LLM (K=5)Refinement=Iterative2026.05 | 8.64 | |
| LLaMA-2-13BTraining Phase=SFT2026.05 | 5.55 |