Item Ranking on RecRM-Bench
86.78AccuracyOurs
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Ours2026.05 | 86.78 | 86.32 | 83.67 | |
| LongCat-Flash-ChatThinking=false2026.05 | 62.48 | 54.93 | 82.93 | |
| Deepseek-V3.2Thinking=false2026.05 | 52.32 | 50.57 | 82.53 | |
| Qwen3-MaxThinking=true2026.05 | 50.16 | 50.01 | 78.42 | |
| Qwen3-MaxThinking=false2026.05 | 50.11 | 50.18 | 77.17 | |
| GPT-4.12026.05 | 42.54 | 51.32 | 75.37 | |
| LongCat-Flash-ThinkingThinking=true2026.05 | 34.09 | 57.89 | 70 | |
| Deepseek-V3.2Thinking=true2026.05 | 29.55 | 57.04 | 72.42 |