Value annotation on Dataset-2 benchmark leaderboard · SOTA2 Research

Value annotation on Dataset-2

72PrecisionGemini bias-calibrated English-language prompt

Evaluation Results

Method
Gemini bias-calibrated English-language prompt2026.06	72	87	79	9	8
Gemini bias-calibrated English-language prompt2026.06	67	74	71	12	13
Gemini bias-calibrated English-language prompt2026.06	64	96	77	32	7
Gemini bias-calibrated English-language prompt2026.06	61	68	64	12	20
Gemini bias-calibrated English-language prompt2026.06	51	82	63	17	14
Gemini bias-calibrated English-language prompt2026.06	50	56	53	7	44
Gemini bias-calibrated English-language prompt2026.06	49	90	64	24	17
Gemini bias-calibrated English-language prompt2026.06	45	94	61	33	6
Gemini bias-calibrated English-language prompt2026.06	35	27	31	4	59
Gemini bias-calibrated English-language prompt2026.06	29	50	37	17	7