Direct Recommendation on Amazon Toys 5-core filtering (leave-one-out)
52.25H@10VLM2Rec
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| VLM2RecModel Category=Multimodal-based, Backbone=Qwen2.5-VL-3B2026.03 | 52.25 | 64.76 | 35.78 | 38.93 | |
| LLMSFTModel Category=Text-based, Training Strategy=SFT2026.03 | 42.87 | 56.69 | 27.12 | 30.61 | |
| VLMSFT(Ext.)Model Category=Multimodal-based, Fusion Strategy=External, Training Strategy=SFT2026.03 | 41.6 | 58.97 | 22.09 | 26.47 | |
| LLMEMBModel Category=Text-based2026.03 | 39.44 | 52.78 | 24.47 | 27.83 | |
| SASRecModel Category=ID-based2026.03 | 38.12 | 47.8 | 26.13 | 28.56 | |
| GRU4RecModel Category=ID-based2026.03 | 35.91 | 45.96 | 22.95 | 25.48 | |
| VLMSFT(Int.)Model Category=Multimodal-based, Fusion Strategy=Internal, Training Strategy=SFT2026.03 | 33.81 | 51.31 | 17.7 | 22.09 | |
| CLIPModel Category=Multimodal-based2026.03 | 32.58 | 43.42 | 20.69 | 23.41 | |
| VLMVanilla(Int.)Model Category=Multimodal-based, Fusion Strategy=Internal2026.03 | 31.84 | 42.45 | 20.93 | 23.6 | |
| LLM2RecModel Category=Text-based2026.03 | 31.72 | 47.18 | 18.75 | 22.62 | |
| VLMPromptModel Category=Multimodal-based2026.03 | 28.92 | 42.05 | 17.17 | 20.46 | |
| SLIM+Model Category=Text-based2026.03 | 28.39 | 42.08 | 16.25 | 19.68 | |
| VLMVanilla(Ext.)Model Category=Multimodal-based, Fusion Strategy=External2026.03 | 26.59 | 42.22 | 14.26 | 18.17 | |
| LLMVanillaModel Category=Text-based, Training Strategy=Vanilla2026.03 | 24.19 | 35.97 | 14.52 | 17.47 | |
| NoteLLM-2Model Category=Multimodal-based2026.03 | 22.37 | 45.45 | 10.03 | 15.82 | |
| BERTModel Category=Text-based2026.03 | 12.42 | 23.36 | 5.79 | 8.53 | |
| SLIMModel Category=Text-based2026.03 | 12.17 | 24.74 | 5.67 | 8.8 |