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

Diagnosis on DermoBench Out-of-Distribution

94Derm1M Score

Human Performance

Updated 1mo ago

Evaluation Results

Method	Links
Human Performance 2026.01		94	86	89	93	90.5
DermoGPT-RL + CCT 2026.01		71.56	62.96	70.13	61.25	66.48
DermoGPT-SFT + CCT 2026.01		70.65	64.33	65.58	61.25	65.45
DermoGPT-RL 2026.01		69.68	62.8	68.59	60	65.27
DermoGPT-SFT 2026.01		68.91	62.8	65.88	59.17	64.19
Gemini-2.5-Flash 2026.01		66.33	59.15	53.96	65.42	61.21
QVQ-72B-Preview 2026.01		60.53	53.66	56.92	62.92	58.51
Qwen2.5-VL-72B 2026.01		54.63	54.88	58.36	66.67	58.63
Claude-Sonnet-4.5-Thinking 2026.01		53.64	52.9	50.4	68.75	56.42
Qwen3-VL-8B (Base) 2026.01		52.67	51.07	59.1	55.42	54.31
GPT-4o-mini 2026.01		52.12	58.54	56.48	59.17	56.57
SkinVL-PubMM 2026.01		51.12	48.93	58.95	54.58	53.4
Qwen3-VL-32B 2026.01		48.13	57.93	63.11	69.58	59.69
GLM-4.5V 2026.01		45.51	48.17	43.08	57.08	48.46
Llama-3.2-90B 2026.01		44.76	49.09	37.14	49.58	45.14
Nemotron-Nano 2026.01		44.12	39.48	36.84	52.08	43.14
LLaVA-Med-v1.5 2026.01		41.38	36.74	33.63	37.08	37.21
HuatuoGPT-Vis-7B 2026.01		31.4	36.13	41.64	47.92	39.27
Lingshu-32B 2026.01		30.29	34.91	32.24	45.83	35.82
Lingshu-7B 2026.01		25.95	32.16	33.88	40	33
Llama-3.2-11B 2026.01		25.5	21.8	26.9	42.92	29.28