User Representation Ranking on Alipay 10 Scenarios (test)
0.2412ActiveQ-Anchor (Prompt Tuned)
Evaluation results
| Method | Links | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Q-Anchor (Prompt Tuned)Model Category=Q-Anchor Embedding, Training Stage=post-trained, Prompt Tuning=true2026.02 | 0.2412 | 0.1177 | 0.5225 | 0.848 | 0.7086 | 0.7573 | 0.6212 | 0.5564 | 0.1962 | 0.6978 | 0.5267 | |
| One4allModel Category=User Embedding Models2026.02 | 0.2403 | 0.1075 | 0.7749 | 0.8191 | 0.716 | 0.6113 | 0.3886 | 0.2135 | 0.1261 | 0.498 | 0.4495 | |
| CPCModel Category=User Embedding Models2026.02 | 0.2402 | 0.0921 | 0.7729 | 0.8197 | 0.7077 | 0.6028 | 0.4249 | 0.2068 | 0.1824 | 0.5061 | 0.4556 | |
| Q-Anchor (Base)Model Category=Q-Anchor Embedding, Training Stage=pretrained2026.02 | 0.2228 | 0.1023 | 0.5216 | 0.8437 | 0.7054 | 0.745 | 0.621 | 0.4527 | 0.1727 | 0.6567 | 0.5044 | |
| MSDPModel Category=User Embedding Models2026.02 | 0.2165 | 0.0608 | 0.7731 | 0.8104 | 0.7026 | 0.6069 | 0.4229 | 0.2107 | 0.1299 | 0.5345 | 0.4469 | |
| FOUNDModel Category=User Embedding Models2026.02 | 0.1669 | 0.078 | 0.466 | 0.7967 | 0.6739 | 0.7084 | 0.5762 | 0.3348 | 0.1753 | 0.5521 | 0.4529 | |
| Llama-Embed-Nemotron-8BModel Category=General Embedding Models, Model Scale=8B2026.02 | 0.1302 | 0.0857 | 0.4003 | 0.5201 | 0.5758 | 0.5868 | 0.4918 | 0.3685 | 0.1364 | 0.5091 | 0.3805 | |
| KaLM-Embed.-Gemma3-12BModel Category=General Embedding Models, Model Scale=12B2026.02 | 0.088 | 0.0533 | 0.383 | 0.4036 | 0.4944 | 0.4879 | 0.4207 | 0.2139 | 0.1312 | 0.3937 | 0.307 | |
| Qwen3-Embedding-0.6BModel Category=General Embedding Models, Model Scale=0.6B2026.02 | 0.0524 | 0.033 | 0.3459 | 0.3274 | 0.3102 | 0.3086 | 0.275 | 0.1581 | 0.0814 | 0.327 | 0.2219 | |
| Qwen2.5-0.5B-InstructModel Category=General Embedding Models, Model Scale=0.5B2026.02 | 0.0421 | 0.028 | 0.3326 | 0.3174 | 0.3034 | 0.2958 | 0.2066 | 0.1324 | 0.088 | 0.3262 | 0.2073 |