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

Multi-modal Reasoning on CV-Bench

86.5Overall Accuracy

Robix 7B Base

67.384872.347477.3182.2726Dec 9, 2025Jan 8, 2026Feb 8, 2026Mar 11, 2026Apr 11, 2026May 12, 2026Jun 12, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2025.12
86.5-
2025.12
86.36-
2025.12
85.81-
2025.12
84.93-
2025.12
81.78-
2025.12
79.1-
2026.06
78.247.36
2026.06
78.098.64
2026.06
77.758.08
2026.06
77.529.56
2026.06
76.59.11
2026.06
76.0410.43
2026.06
68.3912.44
2026.06
68.1217.77