Soft prompt threats: Attacking safety alignment and unlearning in open-source llms through the embedding space
Current research in adversarial robustness of LLMs focuses on discrete input manipulations
in the natural language space, which can be directly transferred to closed-source models …
in the natural language space, which can be directly transferred to closed-source models …
Leveraging Natural Language Processing for In-Depth Analysis and Insights from Hospital Patient Feedback Data
N Taj, GS Girisha - 2024 Third International Conference on …, 2024 - ieeexplore.ieee.org
The patient's feedback is a key component in enhancing healthcare service. This research
employed sophisticated NLP tools to analyze the feedback of 500 hospitals in Bangalore …
employed sophisticated NLP tools to analyze the feedback of 500 hospitals in Bangalore …