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

Embodied Question Answering on OpenEQA

69.3Accuracy

GPT-5.4

22.39634.57346.7558.927May 4, 2026May 10, 2026May 16, 2026May 22, 2026May 28, 2026Jun 3, 2026Jun 9, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.06
69.3-
2026.06
65.8-
2026.06
65.1-
2026.05
64.4-
2026.06
62.6-
2026.05
59.2-
2026.05
55-
2026.05
55-
2026.06
54.7-
2026.06
53.6-
2026.06
51.1-
2026.05
50.5-
2026.06
50.5-
2026.05
49.3-
2026.05
47.6-
2026.05
46.9-
2026.05
44.7-
2026.05
44.6-
2026.05
43.5-
2026.06
42.9-
2026.05
36.2-
2026.06
35.1-
2026.06
34.5-
2026.06
32.7-
2026.06
27.8-
2026.06
24.2-
2026.01
-54.4
2026.01
-55.5
2026.01
-54
2026.01
-50.8
2026.01
-49.1
2026.01
-51.1
2026.01
-56.2
2026.01
-59
2026.03
-56.4
2026.03
-59.9
2026.03
-68.3
2026.03
-64.9
2026.03
-56.4
2026.03
-67.1
2026.03
-60
2026.03
-62.6
2026.03
-63.8
2026.03
-63.3
2026.03
-74.1
2026.03
-56.3
2026.03
-70