Terminology Understanding on Code Domain Dataset
98Recall@10LBR
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| LBRBackbone=Qwen2.5-1.5B, Attention Architecture=Causal, Training Paradigm=LLM+IB-GL+CL2026.01 | 98 | 88.7 | 87.91 | |
| LBRBackbone=Llama3.2-1B, Attention Architecture=Causal, Training Paradigm=LLM+IB-GL+CL2026.01 | 97.9 | 87.9 | 84.53 | |
| LBRBackbone=Qwen2-1.5B, Attention Architecture=Causal, Training Paradigm=LLM+IB-GL+CL2026.01 | 97.8 | 88.2 | 87.82 | |
| SFT+CLBackbone=Qwen2-1.5B, Attention Architecture=Causal, Training Paradigm=LLM+GL+CL2026.01 | 88.8 | 76.4 | 64.47 | |
| BGEBackbone=XLMR-Large, Attention Architecture=Bi., Training Paradigm=LLM+CL2026.01 | 88.2 | 75.4 | 78.35 | |
| SFT+CLBackbone=Qwen2.5-1.5B, Attention Architecture=Causal, Training Paradigm=LLM+GL+CL2026.01 | 85.5 | 73.2 | 60.68 | |
| LLM2VecBackbone=SLLaMA-1.3B, Attention Architecture=Bi., Training Paradigm=LLM+CL2026.01 | 85 | 72.9 | 79.34 | |
| GTEBackbone=Qwen2-1.5B, Attention Architecture=Bi., Training Paradigm=LLM+CL2026.01 | 36 | 28.6 | 65.26 | |
| Qwen2Backbone=Qwen2-1.5B, Attention Architecture=Causal, Training Paradigm=LLM+GL2026.01 | 15.1 | 10.8 | 20.01 | |
| HuaTuoBackbone=LLaMA-7B, Attention Architecture=Causal, Training Paradigm=LLM+GL2026.01 | 13 | 10.1 | 31.29 | |
| Qwen2.5Backbone=Qwen2.5-1.5B, Attention Architecture=Causal, Training Paradigm=LLM+GL2026.01 | 11.6 | 8.1 | 18.311 | |
| ChemLLMBackbone=InternLM-2B, Attention Architecture=Causal, Training Paradigm=LLM+GL2026.01 | 11.6 | 8.2 | 11.012 |