Zero-shot Reasoning on Downstream Tasks (LMB, PIQA, HellaSwag, OPQA, ARC) benchmark leaderboard · SOTA2 Research

Zero-shot Reasoning on Downstream Tasks (LMB, PIQA, HellaSwag, OPQA, ARC)

32.75LAMBADA (LMB) AccuracyFLT

Evaluation Results

Method	Links
FLT2026.05		32.75	65.07	37.05	33.2	53.87	28.41	41.72
FLT2026.05		31.15	64.85	37.03	35.2	52.81	27.24	41.38
LTi2026.05		31.12	64.68	35.79	33.2	56.06	28.41	41.54
FLT2026.05		30.64	63.81	35.72	31.8	51.09	26.36	39.9
LTi2026.05		30.56	64.85	36.24	32.8	53.78	27.81	41
LTai2026.05		30.35	62.89	36.33	34	52.81	26.62	40.05
LT2026.05		30.07	64.25	35.43	34.6	51.64	27.13	40.52
LTi2026.05		29.92	64.63	36.33	31.4	52.65	26.27	40.2
LTai2026.05		29.53	62.94	35.81	32.4	51.97	27.64	40.04
LTai2026.05		28.39	64.41	35.94	32	50.96	28.07	39.96
LT2026.05		26.37	61.31	31.36	32.8	43.22	24.31	36.56
LTi2026.05		20.58	58.65	27.76	28	38.84	24.23	33.01
LTi2026.05		20.39	59.24	27.71	29.39	39.94	23.46	33.35
LTi2026.05		20.04	59.63	27.8	27.2	39.56	22.44	32.77
LTai2026.05		19.75	58.86	27.5	28.2	39.6	23.54	32.9
LTai2026.05		19.65	59.84	27.56	28.2	39.68	24.65	33.26
FLT2026.05		19.25	58.97	27.91	27.8	37.45	23.03	32.4
FLT2026.05		19.03	59.14	28.2	25.6	38.88	23.89	32.45
LT2026.05		18.7	59.46	27.66	27	38.97	23.63	32.57
FLT2026.05		18.39	59.14	27.48	25.8	39.56	22.44	32.13
LT2026.05		16.33	57.56	26.79	28.2	36.65	23.29	31.47
LT2026.05		12.32	56.9	26	26.8	36.48	22.86	30.22