SERE: Boosting LLMs for Accurate Event Causality Detection
Discover how SERE enhances large language models with structural example retrieval to improve event causality identification accuracy in NLP tasks.
LLM-Based Smart Contract Vulnerability Detection Framework
Discover a new LLM framework for smart contract security with high accuracy in detecting vulnerabilities across multiple blockchain platforms.
Hierarchy-Aware GNN Embeddings for Yeast Phenotype Prediction
Discover how hierarchy-aware graph neural network embeddings improve yeast phenotype prediction by leveraging knowledge graphs and semantic loss.
FUS3DMaps: Scalable Open-Vocabulary 3D Semantic Mapping
Discover FUS3DMaps, a scalable and accurate 3D semantic mapping method using voxel and instance-level fusion for open-vocabulary robotic perception.
ELAS: Efficient Low-Rank LLM Pre-Training with 2:4 Sparsity
Discover ELAS, a framework that boosts low-rank LLM pre-training efficiency using 2:4 activation sparsity, reducing memory and speeding up training.