Elisabeth Kirsten

Bochum, Germany

Google Scholar GitHub LinkedIn

👋 Short bio. I am a PhD student at Ruhr University Bochum and Research Center for Trustworthy Data Science and Security. I am coadvised by Dr. Muhammad Bilal Zafar and Dr. Nicole Krämer.

My research focuses on human-centric and responsible Large Language Models. I am passionate about societal impacts of AI systems as they scale across capabilities, domains, and human populations. I am dedicated to computer science, ethics, and sociology, addressing a range of issues such as AI alignment, safety, and fairness from a multidisciplinary perspective. I did my MSc at Technische Universität Darmstadt, and interned at EPFL and MPI-SP.

📩 You can reach me at first.last[at]rub.de

news

Apr 28, 2025	Going to NAACL 2025 in New Mexico to present our recent work on “The Impact of Inference Acceleration on Bias in LLMs”
Nov 29, 2024	Gave a talk “On Trustworthiness of Large Language Models” at the NET-hUmAINE Winter School `Slides` `Code`

selected publications

NAACL

The Impact of Inference Acceleration on Bias of LLMs

Elisabeth Kirsten, Ivan Habernal, Vedant Nanda, and 1 more author

In Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers), Apr 2025

Abs PDF Code

Last few years have seen unprecedented advances in capabilities of Large Language Models (LLMs). These advancements promise to benefit a vast array of application domains. However, due to their immense size, performing inference with LLMs is both costly and slow. Consequently, a plethora of recent work has proposed strategies to enhance inference efficiency, e.g., quantization, pruning, and caching. These acceleration strategies reduce the inference cost and latency, often by several factors, while maintaining much of the predictive performance measured via common benchmarks. In this work, we explore another critical aspect of LLM performance: demographic bias in model generations due to inference acceleration optimizations. Using a wide range of metrics, we probe bias in model outputs from a number of angles. Analysis of outputs before and after inference acceleration shows significant change in bias. Worryingly, these bias effects are complex and unpredictable. A combination of an acceleration strategy and bias type may show little bias change in one model but may lead to a large effect in another. Our results highlight a need for in-depth and case-by-case evaluation of model bias after it has been modified to accelerate inference.This paper contains prompts and outputs which may be deemed offensive.