Search papers, labs, and topics across Lattice.
Affiliation:
3
0
3
0
The unexpected effects of visual cues on LVLMs highlight the need to understand the presence of an image and its framing when evaluating the robustness of LVLMs.
TRACE significantly improves multimodal time series representation by effectively inferring missing data, outperforming traditional methods under severe modality missingness.
LVLMs leak visual text style into semantic inference, meaning the font of a word can change the attributes a model associates with the concept it represents.
