SRA achieves 99.71% average attack success across 26 LLMs by optimizing for coherent malicious semantics via the SRHS algorithm, with claimed theoretical guarantees on convergence and transfer.
Large language models in medicine
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
roles
background 1polarities
background 1representative citing papers
LLMs tasked with allocating childhood lead testing resources in Chicago, New York, and DC overlooked high-prevalence neighborhoods and reached only 0.46 average accuracy despite marketed research capabilities.
A hybrid ML-LLM pipeline predicts lung cancer pain episodes at 48h and 72h with accuracies 0.876 and 0.917 on 266 patients, boosting sensitivity by ~10.6-10.7% via LLM augmentation of medication trends and clinical notes.
citing papers explorer
-
LLM-Agnostic Semantic Representation Attack
SRA achieves 99.71% average attack success across 26 LLMs by optimizing for coherent malicious semantics via the SRHS algorithm, with claimed theoretical guarantees on convergence and transfer.
-
Can LLMs Help Allocate Public Health Resources? A Case Study on Childhood Lead Testing
LLMs tasked with allocating childhood lead testing resources in Chicago, New York, and DC overlooked high-prevalence neighborhoods and reached only 0.46 average accuracy despite marketed research capabilities.
-
AI-Driven Prediction of Cancer Pain Episodes: A Hybrid Decision Support Approach
A hybrid ML-LLM pipeline predicts lung cancer pain episodes at 48h and 72h with accuracies 0.876 and 0.917 on 266 patients, boosting sensitivity by ~10.6-10.7% via LLM augmentation of medication trends and clinical notes.