Sequential Recommendation on Instruments (test)
32NDCG@1RecNet
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| RecNetBackbone=Deepseek-R1-Distill-Qwen-32B, embedding dimension=1282026.01 | 32 | 55.58 | 64.17 | |
| KGLAAugmentation=knowledge graphs2026.01 | 26 | 47.62 | 57.42 | |
| AgentCF2026.01 | 25 | 45.24 | 54.62 | |
| SASRec-fullmode=full datasets2026.01 | 23 | 45.19 | 58.86 | |
| AgentCF++Architecture=dual-layer memory2026.01 | 23 | 46.06 | 55.97 | |
| LLMRankmode=zero-shot ranker, Backbone=ChatGPT2026.01 | 21.33 | 42.78 | 50.11 | |
| SASRec-samplemode=sampled datasets2026.01 | 18 | 40.37 | 47.85 | |
| BPR-fullmode=full datasets2026.01 | 16 | 36.22 | 43.8 | |
| BPR-samplemode=sampled datasets2026.01 | 14 | 31.78 | 35.84 | |
| Pop2026.01 | 11 | 21.46 | 42.75 |