List your company

LLM-judge preference scoring on PowerSeeking

3.66Concept ScoreSPLIT
1.79842.28172.7653.2483Feb 2, 2026

Evaluation results

MethodLinks
SPLITModel=Qwen-2.5-7B-IT, Form=Local Weight2026.023.66
SPLITModel=Qwen-2.5-7B-IT, Form=Vector2026.023.65
REPSModel=Qwen-2.5-7B-IT, Form=Vector2026.023.63
SPLITModel=Gemma-2-9B-IT, Form=Vector2026.023.62
REPSModel=Gemma-2-9B-IT, Form=Vector2026.023.61
SPLITModel=Gemma-2-9B-IT, Form=Local Weight2026.023.59
SPLITModel=Qwen-2.5-7B-IT, Form=LoRA2026.023.59
SFTModel=Qwen-2.5-7B-IT, Form=Vector2026.023.58
SPLITModel=Gemma-2-9B-IT, Form=LoRA2026.023.56
SFTModel=Qwen-2.5-7B-IT, Form=Local Weight2026.023.53
SFTModel=Gemma-2-9B-IT, Form=Local Weight2026.023.5
REPSModel=Gemma-2-9B-IT, Form=LoRA2026.023.44
SFTModel=Gemma-2-9B-IT, Form=LoRA2026.023.41
REPSModel=Gemma-2-9B-IT, Form=Local Weight2026.023.39
REPSModel=Qwen-2.5-7B-IT, Form=LoRA2026.023.34
SFTModel=Gemma-2-9B-IT, Form=Vector2026.023.3
REPSModel=Qwen-2.5-7B-IT, Form=Local Weight2026.023.24
DiffMeanModel=Qwen-2.5-7B-IT, Form=Vector2026.023.17
SFTModel=Qwen-2.5-7B-IT, Form=LoRA2026.023.05
DiffMeanModel=Gemma-2-9B-IT, Form=Vector2026.022.95
VanillaModel=Qwen-2.5-7B-IT, Form=Vanilla2026.022.24
VanillaModel=Gemma-2-9B-IT, Form=Vanilla2026.021.87