Recommendation on Yelp InstructRec retrospective (test)
58.7HR@1ARS
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| ARSEvaluation Mode=Retrospective, Backbone LLM=Llama-4-Maverick-17B-128E-Instruct, History Window (Lmax)=152026.05 | 58.7 | 88.7 | 75.2 | 78.7 | |
| MemRecEvaluation Mode=Retrospective, Backbone LLM=Llama-4-Maverick-17B-128E-Instruct, History Window (Lmax)=152026.05 | 40.2 | 75.7 | 59 | 66.5 | |
| LightGCNEvaluation Mode=Retrospective2026.05 | 31.3 | 81.7 | 56.2 | 62.4 | |
| iAgentEvaluation Mode=Retrospective, Backbone LLM=Llama-4-Maverick-17B-128E-Instruct, History Window (Lmax)=152026.05 | 28.9 | 68.1 | 48.1 | 58.2 | |
| LLMRankEvaluation Mode=Retrospective, Backbone LLM=Llama-4-Maverick-17B-128E-Instruct2026.05 | 26.5 | 52 | 39.8 | 54.6 | |
| i2AgentEvaluation Mode=Retrospective, Backbone LLM=Llama-4-Maverick-17B-128E-Instruct, History Window (Lmax)=152026.05 | 25.9 | 67.3 | 46.8 | 56.9 | |
| BPREvaluation Mode=Retrospective2026.05 | 23.1 | 66.6 | 44.4 | 55.3 | |
| SASRecEvaluation Mode=Retrospective2026.05 | 19.8 | 45.4 | 33.2 | 49.7 | |
| AgentCFEvaluation Mode=Retrospective, Backbone LLM=Llama-4-Maverick-17B-128E-Instruct2026.05 | 15 | 47.9 | 31.5 | 47.4 |