| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Preference-aligned Retrieval-Augmented Generation | PrefEval | Accuracy77.96 | 27 | |
| Question Answering | PrefEval | Accuracy99.99 | 12 | |
| Personalization Evaluation | PrefEval 10 injected adversarial turns | Pref Unaware Rate7.4 | 10 | |
| Preference evaluation via multi-choice queries | PrefEval Implicit | Accuracy69.9 | 8 | |
| Preference evaluation via multi-choice queries | PrefEval Explicit | Accuracy81.3 | 8 | |
| LLM Preference Alignment | PrefEval | AccPF68.8 | 7 | |
| Preference Alignment | PrefEval Implicit Preference (test) | Score19.1 | 6 | |
| Preference Alignment | PrefEval Implicit Choice (test) | Choice Accuracy78.6 | 6 | |
| Preference Alignment | PrefEval Explicit Preference (test) | LLM-Evaluated Score77.7 | 6 | |
| Personalization | PrefEval (test) | Accuracy38.33 | 4 |