Personalized Recommendation on RecBench+ Book
67.1Performance (Simple)ReRec
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| ReRecCategory=RFT-trained Model, Base Model=Llama-3.2-3B-Instruct2026.04 | 67.1 | 78.2 | 75 | 81.1 | — | |
| GRPOCategory=RFT-trained Model, Base Model=Llama-3.2-3B-Instruct2026.04 | 66.4 | 75.7 | 71.3 | 78.6 | — | |
| REINFORCE++Category=RFT-trained Model, Base Model=Llama-3.2-3B-Instruct2026.04 | 66.1 | 76.8 | 69.7 | 79.5 | — | |
| RLOOCategory=RFT-trained Model, Base Model=Llama-3.2-3B-Instruct2026.04 | 66 | 77.4 | 70.4 | 79.4 | — | |
| RLOOCategory=RFT-trained Model, Base Model=Qwen-2.5-3B-Instruct2026.04 | 56.7 | 64.9 | 53.2 | 71.6 | — | |
| TallRecCategory=LLM-based CRS2026.04 | 56.3 | 59.1 | 25.1 | 47.7 | — | |
| GRPOCategory=RFT-trained Model, Base Model=Qwen-2.5-3B-Instruct2026.04 | 56.3 | 63 | 55.2 | 69.9 | — | |
| Deepseek-R1Category=LLM Backbone2026.04 | 56.2 | 53 | 27.9 | 50.5 | — | |
| CRAGCategory=LLM-based CRS2026.04 | 56 | 59.2 | 21.1 | 51.8 | — | |
| InteRecAgentCategory=LLM-based CRS2026.04 | 55.7 | 58.2 | 14.7 | 49.3 | — | |
| ReRecCategory=RFT-trained Model, Base Model=Qwen-2.5-3B-Instruct2026.04 | 55.5 | 65.5 | 56.2 | 74.6 | — | |
| GPT-4oCategory=LLM Backbone2026.04 | 55.4 | 59 | 16 | 45.8 | — | |
| REINFORCE++Category=RFT-trained Model, Base Model=Qwen-2.5-3B-Instruct2026.04 | 55.3 | 61.8 | 51 | 71.6 | — | |
| Qwen-2.5-3B-InstructCategory=LLM Backbone2026.04 | 37.1 | 30.4 | 17.7 | 41.6 | — | |
| Llama-3.2-3B-InstructCategory=LLM Backbone2026.04 | 21.5 | 13.8 | 10.6 | 25.4 | — | |
| DS-R1-Distill-Qwen-7BCategory=LLM Backbone2026.04 | 13.1 | 10.4 | 8.7 | 22.1 | — |