Recommendation on LFM-1K
93.4AccuracyMixRAGRec
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| MixRAGRecBackbone=LLaMA3-8B2026.05 | 93.4 | 99 | 99.2 | |
| MixRAGRecBackbone=Mistral-7B2026.05 | 91.2 | 95.3 | 96.2 | |
| Rec-r1Backbone=LLaMA3-8B2026.05 | 87.2 | 92.6 | 93.8 | |
| K-RagRecBackbone=LLaMA3-8B2026.05 | 84.5 | 88.6 | 90.6 | |
| K-RagRecBackbone=Mistral-7B2026.05 | 81.8 | 87.4 | 89.4 | |
| G-retrieverBackbone=Mistral-7B2026.05 | 76.3 | 85.6 | 89 | |
| Gemini-2.0-FlashEvaluation Protocol=Zero-shot Inference2026.05 | 72.8 | — | — | |
| TallRecBackbone=LLaMA3-8B2026.05 | 69.9 | 84.8 | 89.1 | |
| G-retrieverBackbone=LLaMA3-8B2026.05 | 69 | 82.5 | 86 | |
| GPT-4oEvaluation Protocol=Zero-shot Inference2026.05 | 68.5 | — | — | |
| TallRecBackbone=Mistral-7B2026.05 | 67.4 | 81.2 | 86.7 | |
| Deepseek-r1Evaluation Protocol=Zero-shot Inference2026.05 | 63.4 | — | — | |
| KG-TextBackbone=LLaMA3-8B2026.05 | 54.5 | — | — | |
| KAPINGBackbone=Mistral-7B2026.05 | 53.6 | — | — | |
| KAPINGBackbone=LLaMA3-8B2026.05 | 53 | — | — | |
| KG-TextBackbone=Mistral-7B2026.05 | 49.2 | — | — | |
| Rec-r1Backbone=Mistral-7B2026.05 | 36.2 | 47.7 | 55.2 | |
| Mistral-7BBackbone=Mistral-7B, Evaluation Protocol=Zero-shot Inference2026.05 | 31.8 | — | — | |
| LLaMA3-8BBackbone=LLaMA3-8B, Evaluation Protocol=Zero-shot Inference2026.05 | 23.4 | — | — |