Jane Motz Hayes
@janegrlinc.bsky.social
creative design. UX. film. tech. a11y. basketball. F1. 🇨🇦
Not necessarily in that order.
Not necessarily in that order.
Reposted by Jane Motz Hayes
Here's another example from our lab, just accepted at ICLR, young kids can do these easily but Multi Modal Models fail.
arxiv.org/abs/2407.177...
arxiv.org/abs/2407.177...
KiVA: Kid-inspired Visual Analogies for Testing Large Multimodal Models
This paper investigates visual analogical reasoning in large multimodal models (LMMs) compared to human adults and children. A "visual analogy" is an abstract rule inferred from one image and applied ...
arxiv.org
January 23, 2025 at 6:31 PM
Here's another example from our lab, just accepted at ICLR, young kids can do these easily but Multi Modal Models fail.
arxiv.org/abs/2407.177...
arxiv.org/abs/2407.177...