A training-free intervention that broadcasts the densest vision-sink attention head's map to all heads in shallow layers lowers hallucination rates on CHAIR across several LVLMs.
An image is worth 1/2 tokens after layer 2: Plug-and-play inference acceleration for large vision-language models
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Seeing Clearly by Layer Two: Enhancing Attention Heads to Alleviate Hallucination in LVLMs
A training-free intervention that broadcasts the densest vision-sink attention head's map to all heads in shallow layers lowers hallucination rates on CHAIR across several LVLMs.