ResearchBenchmarksPreference Ranking on UltraFeedback Distinct (test)Follow79Ranking Accuracy (RA)SA-DPO76.50477.15277.878.448Apr 30, 2026Evaluation ResultsMethodMethodLinksRanking Accuracy (RA)SA-DPOBackbone=Llama-3-8B, F...Backbone=Llama-3-8B, Fine-tuning Protocol=LoRA2026.0479SimPOBackbone=Llama-3-8B, F...Backbone=Llama-3-8B, Fine-tuning Protocol=LoRA2026.0478DPOBackbone=Llama-3-8B, F...Backbone=Llama-3-8B, Fine-tuning Protocol=LoRA2026.0476.6