SwimBird: Eliciting Switchable Reasoning Mode in Hybrid Autoregressive MLLMs
Open paper
Benchmark
Model
Rank
Results
visual-question-answering-on-v-bench
SwimBird
#8
Accuracy: 85.5
visual-question-answering-vqa-on-realworldqa
SwimBird
#35
Accuracy: 73.1
vision-language-understanding-on-mmstar
SwimBird
#36
Accuracy: 71.2
Rank counts only results with a code link.