Pith. sign in

Honeybee: Locality-enhanced projector for multimodal llm

1 Pith paper cite this work. Polarity classification is still indexing.

1 Pith paper citing it

fields

cs.CV 1

years

2024 1

verdicts

CONDITIONAL 1

representative citing papers

FoPru: Focal Pruning for Efficient Large Vision-Language Models

cs.CV · 2024-11-21 · conditional · novelty 6.0

Focal Pruning (FoPru) retains as few as 25% of visual tokens, selected by vision-encoder attention, and keeps accuracy within about 1% on three LVLM benchmarks while cutting time-to-first-token by up to 2.5x.

citing papers explorer

Showing 1 of 1 citing paper.

  • FoPru: Focal Pruning for Efficient Large Vision-Language Models cs.CV · 2024-11-21 · conditional · none · ref 6

    Focal Pruning (FoPru) retains as few as 25% of visual tokens, selected by vision-encoder attention, and keeps accuracy within about 1% on three LVLM benchmarks while cutting time-to-first-token by up to 2.5x.