Rethinking Privacy in LLMs with Information Sufficiency
Explore how Information Sufficiency and pseudonymization improve privacy in LLM communications, balancing data protection and message utility.
LLMs for Toxic Habit Detection in Spanish Clinical Texts
Explore how large language models excel in extracting toxic habits from Spanish clinical texts, improving substance use recognition in healthcare.
Universal Foundation Model for Graph-Structured Data
Discover a universal foundation model enabling transferable graph representations for biomedical and network science applications with superior generalizat...
MorphDistill: Enhanced Colorectal Cancer Survival Prediction
MorphDistill improves colorectal cancer survival prediction by unifying pathology models into a compact encoder for accurate prognostic modeling.
Evidential Deep Learning for Uncertainty in Aquatic AVs
Discover how Evidential Deep Learning improves uncertainty estimation and reconstruction accuracy in autonomous vehicles for aquatic disaster response.