Search papers, labs, and topics across Lattice.
2
0
4
Systematic misalignments in MLLM-generated captions can be detected with 63.8% accuracy, revealing a critical flaw in image-text pairing that has been largely overlooked.
Active data curation during pretraining lets you build a chest X-ray foundation model that rivals full-data models using just 23% of the data and compute.
