Sequential Recommendation on Instrument (test)
9.66Hit Rate@5LC-Rec + ILRec
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| LC-Rec + ILRecBackbone LLM=Llama3.1-8B, Optimization Strategy=ILRec2026.02 | 9.66 | 11.43 | 8.32 | 8.89 | |
| LC-Rec + SDPOBackbone LLM=Llama3.1-8B, Optimization Strategy=SDPO2026.02 | 8.94 | 10.69 | 7.81 | 8.36 | |
| LC-RecBackbone LLM=Llama3.1-8B, Optimization Strategy=SFT-based, Ranking Strategy=Semantic IDs2026.02 | 8.88 | 10.62 | 7.76 | 8.32 | |
| LC-Rec + SPRecBackbone LLM=Llama3.1-8B, Optimization Strategy=SPRec self-play fine-tuning2026.02 | 8.88 | 10.41 | 7.75 | 8.25 | |
| LC-Rec + RosePOBackbone LLM=Llama3.1-8B, Optimization Strategy=RosePO preference optimization2026.02 | 8.61 | 10.06 | 7.6 | 8.07 | |
| BIGRec + ILRecBackbone LLM=Llama3.1-8B, Optimization Strategy=ILRec2026.02 | 8.44 | 10.91 | 7.88 | 8.56 | |
| BIGRec + SPRecBackbone LLM=Llama3.1-8B, Optimization Strategy=SPRec self-play fine-tuning2026.02 | 8.01 | 10.21 | 7.51 | 8.08 | |
| BIGRec + SDPOBackbone LLM=Llama3.1-8B, Optimization Strategy=SDPO2026.02 | 7.93 | 10.16 | 7.45 | 8.06 | |
| BIGRecBackbone LLM=Llama3.1-8B, Optimization Strategy=SFT-based2026.02 | 7.86 | 10.04 | 7.42 | 7.99 | |
| BIGRec + RosePOBackbone LLM=Llama3.1-8B, Optimization Strategy=RosePO preference optimization2026.02 | 7.72 | 9.83 | 7.33 | 7.86 | |
| GRU4RecType=Traditional sequential model2026.02 | 6.75 | 7.73 | 5.16 | 5.54 | |
| SASRecType=Traditional sequential model2026.02 | 6.19 | 6.98 | 4.74 | 5.02 | |
| CaserType=Traditional sequential model2026.02 | 5.02 | 5.83 | 2.87 | 3.34 |