Zero-shot Reasoning on (BoolQ, PIQA, HellaSwag, WinoGrande, ARC) benchmark leaderboard · SOTA2 Research

Zero-shot Reasoning on (BoolQ, PIQA, HellaSwag, WinoGrande, ARC)

71.1BoolQ Accuracy (Zero-shot)FP16

Evaluation Results

Method	Links
FP162025.05		71.1	76.88	72.94	67.09	53.58	40.61	63.7
FP162025.05		68.99	79.05	76.62	69.77	57.95	44.2	66.1
MBOK2025.05		66.94	74.97	65.59	61.72	44.82	34.21	58.04
MBOK2025.05		66.8	77.59	73.79	65.27	54.47	41.89	63.3
MBOK2025.05		66.32	75.84	70.24	62.51	50	37.46	60.4
MoS2025.05		66.12	73.72	63.8	58.98	45.71	33.19	57.09
OneBit2025.05		65.66	71.6	60.07	56.91	45.76	31.74	55.29
MBOK2025.05		65.46	75.79	69.59	62.04	49.11	37.8	59.97
MoS2025.05		65.02	71.55	59.41	56.18	41.84	30.03	54.01
OneBit2025.05		63.06	70.4	54.24	56.67	40.82	29.35	52.42
BiLLM2025.05		62.2	62.51	38.05	56.35	40.69	27.73	47.92
PB-LLM2025.05		62.17	52.82	26.87	50.11	26.89	24.31	40.53
BiLLM2025.05		62.14	59.19	35.18	53.11	34.22	26.54	45.06
OmniQuant2025.05		62.05	62.24	50.34	53.2	40.66	29.61	49.68
LLM-QAT2025.05		59.14	50.12	25.1	49.08	26.26	26.96	35.89
OmniQuant2025.05		58.69	56.53	33.87	51.22	33.63	24.32	43.12
OPTQ2025.05		42.97	49.46	26.19	50.28	26.77	28.58	37.38
OPTQ2025.05		40.61	51.74	25.67	51.85	25.46	27.3	37.11
LLM-QAT2025.05		39.85	49.08	24.37	51.38	27.15	24.32	36.03
PB-LLM2025.05		37.82	53.26	28.89	49.48	28.28	23.72	36.91