Share your thoughts, 1 month free Claude Pro on usSee more

Commonsense Reasoning on Commonsense Tasks

81.35BoolQ Accuracy

Dense

Updated 15d ago

Evaluation Results

Method	Links
Dense 2025.05		81.35	80.78	72.61	79.14	77.69	53.33	45	69.99
Dense 2025.05		77.74	79.11	68.97	75.98	74.62	46.25	44.2	66.7
RAP 2025.05		68.84	76.55	66.11	65.55	62.12	39.51	39.6	59.77
RAP 2025.05		62.81	73.99	63.38	65.77	60.35	36.69	36.6	57.08
ShortGPT 2025.05		62.17	60.12	60.38	43.7	41.25	30.12	35	47.53
LLMPruner 2025.05		62.17	65.02	51.93	42.32	41.33	25.5	29.4	45.37
SliceGPT 2025.05		61.99	65.29	58.08	43.43	52.27	32.08	28.99	48.88
SliceGPT 2025.05		61.87	65.94	54.05	42.51	54.37	31.06	27.8	48.23
SliceGPT 2025.05		61.04	54.56	49.49	28.99	30.68	23.46	25.4	39.09
RAP 2025.05		57.16	58.26	53.75	37.81	37.79	26.02	30.2	43
ShortGPT 2025.05		56.82	58.59	54.85	37.71	36.99	30.2	28	43.3
LLMPruner 2025.05		55.96	61.15	50.83	38.09	35.06	27.13	28.4	42.38
FFN-Skip 2025.05		55.77	51.47	50.74	25.9	25.75	25.25	28	37.56
ShortGPT 2025.05		55.57	50.98	50.51	27.83	26.39	27.47	26.6	37.91
RAP 2025.05		52.2	56.69	50.36	31.91	33.54	24.23	27.4	39.48
FFN-Skip 2025.05		51.93	51.95	50.03	26.02	24.03	26.96	28.59	37.08
LLMPruner 2025.05		50.63	68.82	54.14	52.66	49.33	30.2	34.59	48.63
MHA-Drop 2025.05		46.08	54.56	50.83	29.12	28.62	27.47	25.2	37.41
FFN-Skip 2025.05		44.83	50.64	48.6	25.86	25.72	28.92	28.19	36.13
FFN-Skip 2025.05		42.8	49.35	51.22	26.69	27.18	28.49	26.6	36.05
ShortGPT 2025.05		41.44	50.98	50.03	26.74	25.46	25.34	26.8	35.26
SliceGPT 2025.05		40.86	53.86	48.93	27.97	32.07	23.04	25.8	36.08
LLMPruner 2025.05		40.48	50.05	52.33	26.26	26.77	25.94	27	35.55
MHA-Drop 2025.05		37.83	49.89	49.88	25.72	26.05	25.43	26.8	34.51
MHA-Drop 2025.05		37.83	51.9	50.12	26.04	27.86	22.95	25.8	34.64
MHA-Drop 2025.05		37.8	53.92	50.2	26.87	30.47	23.81	27.2	35.75