A hard-negative gradient amplifier improves multimodal contrastive embedding training, achieving 72.5 average on MMEB, but it is a heuristic reweighting rather than a theoretical advance.
Mmsearch: Unveiling the potential of large models as multi-modal search engines
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
-
Improve Multi-Modal Embedding Learning via Explicit Hard Negative Gradient Amplifying
A hard-negative gradient amplifier improves multimodal contrastive embedding training, achieving 72.5 average on MMEB, but it is a heuristic reweighting rather than a theoretical advance.