A learned spatial token fusion plus multi-block features lets a multimodal model use only 25% of its vision tokens while matching or exceeding baseline accuracy on eight benchmarks.
An image is worth 16x16 words: Transformers for image recognition at scale
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
background 1representative citing papers
citing papers explorer
-
Learning Compact Vision Tokens for Efficient Large Multimodal Models
A learned spatial token fusion plus multi-block features lets a multimodal model use only 25% of its vision tokens while matching or exceeding baseline accuracy on eight benchmarks.