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

Image Classification on ImageNet-1k (Specific Metric Set)

63.2Accuracy (Val)

Cluster-aware Upcycling

Updated 3mo ago

Evaluation Results

Method	Links
Cluster-aware Upcycling 2026.04		63.2	55.1	24.1	72.1	46.8	43.5	50.8
Sparse Upcycling 2026.04		63	55.1	23.7	71.2	46.3	42.3	50.3
CLIP-MoE 2026.04		62.9	54.9	24.5	71.6	46.2	43.4	50.6
Dense 2026.04		62.5	54.4	23.5	70.6	45.8	42.5	49.9
Drop-Upcycling 2026.04		62	54.5	22.7	70.8	45.7	42.9	49.8
Dense 2026.04		59.4	51.7	20.1	67.3	42.9	39.4	46.8
Drop-SVD 2026.04		57.4	48.8	14.1	64.6	42	36.1	43.8
Cluster-aware Upcycling 2026.04		57.3	49.2	14	65.2	42.3	36.5	44.1
Cluster-aware Upcycling 2026.04		57.3	49.2	14	65.2	42.3	36.5	44.1
Sparse Upcycling 2026.04		57.1	49.1	13.8	64.3	41.8	36	43.7
Sparse Upcycling 2026.04		57.1	49.1	13.8	64.3	41.8	36	43.7
DeRS-LM 2026.04		56.8	48.6	13.9	64.2	41.1	36.4	43.5
DeRS-LM 2026.04		56.8	48.6	13.9	64.2	41.1	36.4	43.5
Dense 2026.04		56.7	48.5	13.9	64	41.2	36.1	43.4
Drop-Upcycling 2026.04		56.7	48.4	13.1	63.5	41	35.7	43.1
CLIP-MoE 2026.04		56.6	48.1	14.3	64.2	41.4	35.7	43.4
CLIP-MoE 2026.04		56.6	48.1	14.3	64.2	41.4	35.7	43.4
Drop-Upcycling 2026.04		56	47.7	12.9	63.4	40.8	34.3	42.5
Drop-Upcycling 2026.04		56	47.7	12.9	63.4	40.8	34.3	42.5
Drop-Upcycling 2026.04		55.7	47.3	12.4	62.3	40.1	33	41.8
Dense 2026.04		49.6	41.9	9.7	56.7	34.9	31	37.3