Multimodal LLMs act as training-free similarity estimators for instance-level image retrieval by converting next-token probabilities from image-pair prompts into scores, combined with efficient indexing for scalability.
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
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Indexing Multimodal Language Models for Large-scale Image Retrieval
Multimodal LLMs act as training-free similarity estimators for instance-level image retrieval by converting next-token probabilities from image-pair prompts into scores, combined with efficient indexing for scalability.