Scalable Patient-Trial Matching with Lightweight LLM Models
Discover a scalable, efficient framework using lightweight retrieval-augmented generation and LLMs for accurate patient-trial matching in healthcare.
Call-Chain-Aware LLM Test Generation for Java Projects
Discover CAT, a call-chain-aware LLM approach that boosts Java test generation with improved coverage and reliability for complex software systems.
H-Sets: Discovering Feature Interactions in Image Classifiers
H-Sets framework enhances image classifier interpretability by detecting and attributing set-level feature interactions using Hessian-guided methods.
EgoMAGIC Dataset for Medical AI Training and Perception
Explore EgoMAGIC, a large egocentric video dataset with 3,355 labeled medical task videos to train AI perception algorithms in healthcare.
Mochi: Efficient Graph Models via Meta-Learning Alignment
Discover Mochi, a meta-learning framework that aligns pre-training and inference for faster, more efficient graph foundation models with superior performan...