6 papers
Natural-Language-Guided Generator-Agnostic Shortlisting for Protein Binder Design
Gyubok Lee, Kiwoong Yoo, Jimin Seo +2
Modern de novo design workflows generate many candidate protein binders, but wet-lab validation capacity remains limited, making shortlisting a major bottleneck. We study whether L…
LayerNorm Induces Recency Bias in Transformer Decoders
Junu Kim, Xiao Liu, Zhenghao Lin +3
Causal self-attention provides positional information to Transformer decoders. Prior work has shown that stacks of causal self-attention layers alone induce a positional bias in at…
Federated Learning for Heterogeneous Electronic Health Record Systems with Cost Effective Participant Selection
Jiyoun Kim, Junu Kim, Kyunghoon Hur +1
The increasing volume of electronic health records (EHRs) presents the opportunity to improve the accuracy and robustness of models in clinical prediction tasks. Unlike traditional…
MD-ViSCo: A Unified Model for Multi-Directional Vital Sign Waveform Conversion
Franck Meyer, Kyunghoon Hur, Edward Choi
Despite the remarkable progress of deep-learning methods generating a target vital sign waveform from a source vital sign waveform, most existing models are designed exclusively fo…
Towards Predicting Temporal Changes in a Patient's Chest X-ray Images based on Electronic Health Records
Daeun Kyung, Junu Kim, Tackeun Kim +1
Chest X-ray (CXR) is an important diagnostic tool widely used in hospitals to assess patient conditions and monitor changes over time. Recently, generative models, specifically dif…
Enhancing LLMs' Clinical Reasoning with Real-World Data from a Nationwide Sepsis Registry
Junu Kim, Chaeeun Shim, Sungjin Park +16
Although large language models (LLMs) have demonstrated impressive reasoning capabilities across general domains, their effectiveness in real-world clinical practice remains limite…