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

Progress Estimation on PROGRESS-BENCH Answerable samples 1.0

13.8NSE (Vision)

ProgressLM-3B-RL

Updated 1mo ago

Evaluation Results

Method	Links
ProgressLM-3B-RL 2026.01		13.8	90.1	8.5	21.2	63.9	5.6	15.3	84.8	7.9	17.5	77	7
GPT-5 2026.01		18.9	89.4	1.3	23.6	55.8	7	19.9	82.5	2.5	21.3	72.6	4.2
ProgressLM-3B-SFT 2026.01		19	72.4	6.5	29.1	46.3	9.2	21.1	67	7	24	59.3	7.8
Qwen3-VL-32B 2026.01		20.2	80.7	0.3	25.1	51.9	7.4	21.2	74.8	1.7	22.6	66.3	3.9
GPT-5-mini 2026.01		20.7	87.7	0.4	21.1	55.2	9.7	20.8	81.1	2.3	20.9	71.4	5.1
Qwen3-VL-8B 2026.01		23.5	67	0	35.3	34.2	3.4	25.9	60.3	0.7	29.4	50.6	1.7
Qwen2.5-VL-72B 2026.01		27	60	5.9	38.9	41.6	0	29.4	56.2	4.7	32.9	50.8	3
Qwen3-VL-4B 2026.01		30.6	57.4	0	36.2	38.1	0.2	31.8	53.4	0.04	33.4	47.8	0.1
Qwen2.5-VL-3B 2026.01		32.2	32.8	0.02	45.9	7.5	0	35	27.6	0.02	39	20.2	0.01
Qwen2.5-VL-7B 2026.01		34	33.7	28.3	39.1	20.5	0	35	31	22.5	36.5	27.1	14.2
Intern3.5-VL-38B 2026.01		35.2	56.7	37.1	26.5	23.3	43.1	33.4	49.9	38.3	30.8	40	40.1
Qwen2.5-VL-32B 2026.01		38.9	41.5	0	42.6	30	0	39.7	39.2	0	40.8	35.8	0
Intern3.5-VL-4B 2026.01		43.7	18	0.3	37	5.6	10.9	42.3	15.5	2.5	40.4	11.8	5.6
Qwen3-VL-2B 2026.01		64.5	32.1	5.2	59.3	-	12.5	63.4	-	6.7	61.9	-	8.9
Intern3.5-VL-8B 2026.01		64.9	8.2	0.4	33.7	26.7	17.9	58.5	11.9	4	49.3	17.4	9.2
Intern3.5-VL-14B 2026.01		65.2	-22.3	0.2	39.5	10.3	6.8	60	-15.6	1.5	52.3	-6	3.5