Red Teaming Large Language Models for Healthcare
Authors:
Vahid Balazadeh,
Michael Cooper,
David Pellow,
Atousa Assadi,
Jennifer Bell,
Mark Coastworth,
Kaivalya Deshpande,
Jim Fackler,
Gabriel Funingana,
Spencer Gable-Cook,
Anirudh Gangadhar,
Abhishek Jaiswal,
Sumanth Kaja,
Christopher Khoury,
Amrit Krishnan,
Randy Lin,
Kaden McKeen,
Sara Naimimohasses,
Khashayar Namdar,
Aviraj Newatia,
Allan Pang,
Anshul Pattoo,
Sameer Peesapati,
Diana Prepelita,
Bogdana Rakova
, et al. (10 additional authors not shown)
Abstract:
We present the design process and findings of the pre-conference workshop at the Machine Learning for Healthcare Conference (2024) entitled Red Teaming Large Language Models for Healthcare, which took place on August 15, 2024. Conference participants, comprising a mix of computational and clinical expertise, attempted to discover vulnerabilities -- realistic clinical prompts for which a large lang…
▽ More
We present the design process and findings of the pre-conference workshop at the Machine Learning for Healthcare Conference (2024) entitled Red Teaming Large Language Models for Healthcare, which took place on August 15, 2024. Conference participants, comprising a mix of computational and clinical expertise, attempted to discover vulnerabilities -- realistic clinical prompts for which a large language model (LLM) outputs a response that could cause clinical harm. Red-teaming with clinicians enables the identification of LLM vulnerabilities that may not be recognised by LLM developers lacking clinical expertise. We report the vulnerabilities found, categorise them, and present the results of a replication study assessing the vulnerabilities across all LLMs provided.
△ Less
Submitted 1 May, 2025;
originally announced May 2025.
Opinion dynamics and the unpredictability of opinion trajectories in an adaptive social network model
Authors:
Akshay Gangadhar,
Hiroki Sayama
Abstract:
Understanding opinion dynamics in social networks is critical for predicting social behavior and detecting polarization. Traditional approaches often rely on static snapshots of network states, which can obscure the underlying dynamics of opinion evolution. In this study, we introduce a dynamic framework that quantifies the unpredictability of opinion trajectories using the normalized Lempel-Ziv (…
▽ More
Understanding opinion dynamics in social networks is critical for predicting social behavior and detecting polarization. Traditional approaches often rely on static snapshots of network states, which can obscure the underlying dynamics of opinion evolution. In this study, we introduce a dynamic framework that quantifies the unpredictability of opinion trajectories using the normalized Lempel-Ziv (nLZ) complexity. Our approach leverages an adaptive social network model where each node is characterized by three behavioral parameters - homophily, neophily, and social conformity - and where opinions evolve continuously according to a system of ordinary differential equations. The results reveal distinct nLZ complexity signatures for each node type: homophilic nodes exhibit consistently rising complexity, reflecting increasingly unpredictable opinion shifts that are counterintuitive given their tendency for similarity; neophilic nodes maintain low and stable complexity, suggesting that openness to novelty can, surprisingly, lead to stable opinion dynamics; and conformic nodes display a U-shaped complexity trend, transitioning from early opinion stagnation to later unpredictability. In fully heterogeneous networks, modest interaction effects emerge, with slight shifts in the unpredictability of each faction's trajectories. These findings underscore the importance of temporal analysis in uncovering hidden dynamical patterns, offering novel insights into the mechanisms underlying social adaptation and polarization.
△ Less
Submitted 10 April, 2025;
originally announced April 2025.
