Model
GraphVis
A vision-language model with approximately 7 billion parameters designed for knowledge-graph question answering. It performs knowledge-graph-augmented question answering by visualizing graph structures as images.
Timeline (last 90 days)
Sep 18
VisKG-LM, using only approximately 400M active parameters, matches or surpasses the 7B-parameter vision-language model GraphVis.
Unverified50%