Cross-domain recommendation on Cross-Live
21.1Pass@64OneReason RFT thinking
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| OneReason RFT thinkingCategory=Itemic Token-Based2026.06 | 21.1 | 18.35 | |
| OneReason RFT non-thinkingCategory=Itemic Token-Based2026.06 | 21.01 | 18.17 | |
| LC-Rec-PT-SFT-8BCategory=Itemic Token-Based2026.06 | 19.32 | 16.7 | |
| OneReason SFT non-thinkingCategory=Itemic Token-Based2026.06 | 18.05 | 15.52 | |
| OneReason SFT thinkingCategory=Itemic Token-Based2026.06 | 16.43 | 14.32 | |
| GPT-5.4Category=Text-Based2026.06 | 7.2 | 6.38 | |
| Qwen3-32BCategory=Text-Based2026.06 | 5.64 | 5.1 | |
| Qwen3-235B-A22BCategory=Text-Based2026.06 | 5.1 | 4.66 | |
| GPT-4o-miniCategory=Text-Based2026.06 | 4.01 | 3.57 | |
| Gemini-3-PreviewCategory=Text-Based2026.06 | 3.92 | 3.44 | |
| LC-Rec-SFT-Only-14BCategory=Itemic Token-Based2026.06 | 3.76 | 3.09 | |
| Deepseek-V3.2Category=Text-Based2026.06 | 3.46 | 3.12 | |
| Claude-Opus-4.6Category=Text-Based2026.06 | 3.02 | 2.66 | |
| HSTUCategory=ID-Based2026.06 | 2.32 | 2.14 | |
| TIGERCategory=Itemic Token-Based2026.06 | 2.32 | 1.78 | |
| Qwen3-8BCategory=Text-Based2026.06 | 2.1 | 1.85 | |
| SASRecCategory=ID-Based2026.06 | 1.76 | 0.4 | |
| LC-Rec-SFT-Only-8BCategory=Itemic Token-Based2026.06 | 0.89 | 0.71 |