Natural Language Understanding on (Winogrande, PIQA, WSC, WNLI, SST-2, RTE, QNLI, CB, ARC-e, ARC-c) (test) benchmark leaderboard · SOTA2 Research

Natural Language Understanding on (Winogrande, PIQA, WSC, WNLI, SST-2, RTE, QNLI, CB, ARC-e, ARC-c) (test)

75.61Winogrande AccuracyDense

Evaluation Results

Method	Links
Dense2026.01		75.61	79.71	53.85	66.2	87.61	69.31	58.56	80.36	81.82	53.16	70.62
Dense2026.01		75.3	80.01	69.23	73.24	95.07	80.51	74.81	87.5	81.14	52.47	76.93
Dense2026.01		72.85	78.24	54.81	59.15	85.89	66.06	54.6	60.71	79	47.78	65.91
Dense2026.01		71.98	79.22	73.08	64.79	94.15	83.75	74.24	76.79	79.04	48.04	74.51
LLM-Pruner2026.01		71.74	73.07	36.54	43.66	66.86	52.71	49.9	42.86	66.08	32.42	53.58
SliceGPT2026.01		70.72	63.17	44.23	43.66	51.15	52.71	50.58	41.07	65.03	33.79	51.61
MI-PRUN2026.01		69.69	73.5	63.46	60.56	83.94	60.29	50.91	62.5	67.21	39.42	63.15
MI-PRUN2026.01		69.46	72.25	81.73	71.83	95.41	80.87	69.89	83.93	67.38	43.69	73.64
MI-PRUN2026.01		69.3	73.78	63.46	61.97	94.04	71.84	50.54	62.5	66.25	39.93	65.36
ShortGPT2026.01		67.72	71.6	36.54	43.66	49.2	53.43	50.19	39.29	65.66	39.42	51.67
ShortGPT2026.01		67.4	73.01	63.46	63.38	93.12	80.87	50.54	60.71	62.96	35.84	65.13
ShortGPT2026.01		67.4	70.18	36.54	71.83	94.15	80.51	49.46	87.5	63.72	39.33	66.06
LLM-Pruner2026.01		64.72	76.06	36.54	56.34	65.83	52.71	51.2	46.43	64.86	36.69	55.14
ShortGPT2026.01		64.17	71.93	58.65	59.15	50.57	68.95	49.83	57.14	50.17	40.53	57.11
SliceGPT2026.01		63.06	67.03	63.46	45.07	52.64	59.57	50.63	42.86	55.26	34.56	53.41
MI-PRUN2026.01		59.59	72.36	59.62	66.2	70.87	69.31	50.61	78.57	60.02	40.27	62.74