Token Activation Maps applied to MLLM art descriptions reveal that visual grounding strength varies by token category, with better artist identification than title prediction.
arXiv preprint arXiv:2602.20853 (2026)
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Understanding How MLLMs Describe Artworks Using Token Activation Maps
Token Activation Maps applied to MLLM art descriptions reveal that visual grounding strength varies by token category, with better artist identification than title prediction.