Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Image Understanding on TextVQA (val)
Loading...
88.39
Accuracy
Vanilla MoE
76.9812
79.9431
82.905
85.8669
Apr 19, 2026
Accuracy
Updated 26d ago
Evaluation Results
Method
Method
Links
Accuracy
Vanilla MoE
Backbone=KIMI-VL-A3B-I...
2026.04
88.39
MACS (Ours)
Backbone=KIMI-VL-A3B-I...
2026.04
88.27
MACS (w/o Expanded)
Backbone=KIMI-VL-A3B-I...
2026.04
87.94
Vanilla MoE
Backbone=INTERNVL3.5-3...
2026.04
85.68
MACS (Ours)
Backbone=INTERNVL3.5-3...
2026.04
85.51
MACS (w/o Expanded)
Backbone=INTERNVL3.5-3...
2026.04
84.92
CAI-MoE (Expanded)
Backbone=KIMI-VL-A3B-I...
2026.04
84.72
Vanilla MoE
Backbone=QWEN3-VL-30B-...
2026.04
83.54
MACS (Ours)
Backbone=QWEN3-VL-30B-...
2026.04
83.41
MACS (w/o Expanded)
Backbone=QWEN3-VL-30B-...
2026.04
83.04
CAI-MoE (Token Drop)
Backbone=KIMI-VL-A3B-I...
2026.04
82.56
CAI-MoE (Expanded)
Backbone=INTERNVL3.5-3...
2026.04
80.94
CAI-MoE (Expanded)
Backbone=QWEN3-VL-30B-...
2026.04
79.86
CAI-MoE (Token Drop)
Backbone=INTERNVL3.5-3...
2026.04
78.13
CAI-MoE (Token Drop)
Backbone=QWEN3-VL-30B-...
2026.04
77.42
Feedback
Search any
task
Search any
task