Product Rating on LaMP-3 1.0 (test)
0.195MAETAP-PER
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| TAP-PERBackbone=Qwen3-4B, variant=+Pu2026.06 | 0.195 | 0.516 | |
| RAGBackbone=Qwen3-4B, k=42026.06 | 0.196 | 0.518 | |
| OPPUBackbone=Qwen3-4B, variant=+rag2026.06 | 0.196 | 0.518 | |
| RAGBackbone=Qwen3-4B, k=12026.06 | 0.223 | 0.543 | |
| RAGBackbone=Qwen3-4B, k=22026.06 | 0.223 | 0.543 | |
| OPPUBackbone=Qwen3-4B, variant=base2026.06 | 0.223 | 0.559 | |
| PAGBackbone=Qwen3-4B, k=02026.06 | 0.241 | 0.559 | |
| OPPUBackbone=Qwen3-4B, variant=+pag2026.06 | 0.241 | 0.559 | |
| PAGBackbone=Qwen3-4B, k=12026.06 | 0.25 | 0.567 | |
| PER-PCSBackbone=Qwen3-4B, variant=+pag2026.06 | 0.259 | 0.59 | |
| TAP-PERBackbone=Qwen3-4B, variant=+Pq2026.06 | 0.268 | 0.582 | |
| P2PBackbone=Qwen3-4B, variant=single2026.06 | 0.275 | 0.6 | |
| PER-PCSBackbone=Qwen3-4B, variant=base2026.06 | 0.277 | 0.605 | |
| PER-PCSBackbone=Qwen3-4B, variant=+rag2026.06 | 0.277 | 0.62 | |
| TAP-PERBackbone=Qwen3-4B, variant=base2026.06 | 0.295 | 0.605 | |
| In-Prompt AugmentationBackbone=FlanT5-XXL, Evaluation Protocol=Zero-shot2023.04 | 0.299 | 0.616 | |
| FlanT5-XXLPersonalization=None, Evaluation Protocol=Zero-shot2023.04 | 0.333 | 0.65 | |
| In-Prompt AugmentationBackbone=GPT-3.5, Evaluation Protocol=Zero-shot2023.04 | 0.603 | 1.002 | |
| GPT-3.5Personalization=None, Evaluation Protocol=Zero-shot2023.04 | 0.677 | 0.948 |