Cross-domain recommendation on Cross-Ad (Pass@64, Recall@64)
17.78Pass@64OneReason RFT thinking
Evaluation results
| Method | Links | ||
|---|---|---|---|
| OneReason RFT thinkingCategory=Itemic Token-Based2026.06 | 17.78 | 7.5 | |
| OneReason RFT non-thinkingCategory=Itemic Token-Based2026.06 | 17.56 | 7.26 | |
| LC-Rec-PT-SFT-8BCategory=Itemic Token-Based2026.06 | 15.85 | 6.55 | |
| OneReason SFT non-thinkingCategory=Itemic Token-Based2026.06 | 15.73 | 6.49 | |
| OneReason SFT thinkingCategory=Itemic Token-Based2026.06 | 9.16 | 3.41 | |
| TIGERCategory=Itemic Token-Based2026.06 | 7.65 | 2.39 | |
| LC-Rec-SFT-Only-14BCategory=Itemic Token-Based2026.06 | 5.99 | 1.94 | |
| LC-Rec-SFT-Only-8BCategory=Itemic Token-Based2026.06 | 2.83 | 0.67 | |
| HSTUCategory=ID-Based2026.06 | 2.79 | 0.78 | |
| GPT-5.4Category=Text-Based2026.06 | 1.64 | 0.43 | |
| Gemini-3-PreviewCategory=Text-Based2026.06 | 1.22 | 0.27 | |
| Qwen3-32BCategory=Text-Based2026.06 | 1.21 | 0.3 | |
| GPT-4o-miniCategory=Text-Based2026.06 | 1.21 | 0.28 | |
| SASRecCategory=ID-Based2026.06 | 1.04 | 0.37 | |
| Qwen3-235B-A22BCategory=Text-Based2026.06 | 0.77 | 0.19 | |
| Deepseek-V3.2Category=Text-Based2026.06 | 0.62 | 0.13 | |
| Claude-Opus-4.6Category=Text-Based2026.06 | 0.5 | 0.11 | |
| Qwen3-8BCategory=Text-Based2026.06 | 0.48 | 0.09 |