Jörg Schlötterer
YOU?
Author Swipe
View article: Breast cancer prediction using mammography exams for real hospital settings
Breast cancer prediction using mammography exams for real hospital settings Open
Breast cancer prediction models for mammography assume that annotations are available for individual images or regions of interest (ROIs), and that there is a fixed number of images per patient. These assumptions do not hold in real hospit…
View article: Marcel: A Lightweight and Open-Source Conversational Agent for University Student Support
Marcel: A Lightweight and Open-Source Conversational Agent for University Student Support Open
We present Marcel, a lightweight and open-source conversational agent designed to support prospective students with admission-related inquiries. The system aims to provide fast and personalized responses, while reducing workload of univers…
View article: Tracing and Reversing Rank-One Model Edits
Tracing and Reversing Rank-One Model Edits Open
Knowledge editing methods (KEs) are a cost-effective way to update the factual content of large language models (LLMs), but they pose a dual-use risk. While KEs are beneficial for updating outdated or incorrect information, they can be exp…
View article: Explanation format does not matter; but explanations do -- An Eggsbert study on explaining Bayesian Optimisation tasks
Explanation format does not matter; but explanations do -- An Eggsbert study on explaining Bayesian Optimisation tasks Open
Bayesian Optimisation (BO) is a family of methods for finding optimal parameters when the underlying function to be optimised is unknown. BO is used, for example, for hyperparameter tuning in machine learning and as an expert support tool …
View article: An XAI-based Analysis of Shortcut Learning in Neural Networks
An XAI-based Analysis of Shortcut Learning in Neural Networks Open
Machine learning models tend to learn spurious features - features that strongly correlate with target labels but are not causal. Existing approaches to mitigate models' dependence on spurious features work in some cases, but fail in other…
View article: Invariant Learning with Annotation-free Environments
Invariant Learning with Annotation-free Environments Open
Invariant learning is a promising approach to improve domain generalization compared to Empirical Risk Minimization (ERM). However, most invariant learning methods rely on the assumption that training examples are pre-partitioned into diff…
View article: Guiding LLMs to Generate High-Fidelity and High-Quality Counterfactual Explanations for Text Classification
Guiding LLMs to Generate High-Fidelity and High-Quality Counterfactual Explanations for Text Classification Open
The need for interpretability in deep learning has driven interest in counterfactual explanations, which identify minimal changes to an instance that change a model's prediction. Current counterfactual (CF) generation methods require task-…
View article: Behavioral Analysis of Information Salience in Large Language Models
Behavioral Analysis of Information Salience in Large Language Models Open
Large Language Models (LLMs) excel at text summarization, a task that requires models to select content based on its importance. However, the exact notion of salience that LLMs have internalized remains unclear. To bridge this gap, we intr…
View article: This looks like what? Challenges and Future Research Directions for Part-Prototype Models
This looks like what? Challenges and Future Research Directions for Part-Prototype Models Open
The growing interest in eXplainable Artificial Intelligence (XAI) has prompted research into models with built-in interpretability, the most prominent of which are part-prototype models. Part-Prototype Models (PPMs) make decisions by compa…
View article: Position: Editing Large Language Models Poses Serious Safety Risks
Position: Editing Large Language Models Poses Serious Safety Risks Open
Large Language Models (LLMs) contain large amounts of facts about the world. These facts can become outdated over time, which has led to the development of knowledge editing methods (KEs) that can change specific facts in LLMs with limited…
View article: Funzac at CoMeDi Shared Task: Modeling Annotator Disagreement from Word-In-Context Perspectives
Funzac at CoMeDi Shared Task: Modeling Annotator Disagreement from Word-In-Context Perspectives Open
In this work, we evaluate annotator disagreement in Word-in-Context (WiC) tasks exploring the relationship between contextual meaning and disagreement as part of the CoMeDi shared task competition. While prior studies have modeled disagree…
View article: Efficient Unsupervised Shortcut Learning Detection and Mitigation in Transformers
Efficient Unsupervised Shortcut Learning Detection and Mitigation in Transformers Open
Shortcut learning, i.e., a model's reliance on undesired features not directly relevant to the task, is a major challenge that severely limits the applications of machine learning algorithms, particularly when deploying them to assist in m…
View article: Enhancing Fact Retrieval in PLMs through Truthfulness
Enhancing Fact Retrieval in PLMs through Truthfulness Open
Pre-trained Language Models (PLMs) encode various facts about the world at their pre-training phase as they are trained to predict the next or missing word in a sentence. There has a been an interest in quantifying and improving the amount…
View article: How to Make LLMs Forget: On Reversing In-Context Knowledge Edits
How to Make LLMs Forget: On Reversing In-Context Knowledge Edits Open
In-context knowledge editing (IKE) enables efficient modification of large language model (LLM) outputs without parameter changes and at zero-cost. However, it can be misused to manipulate responses opaquely, e.g., insert misinformation or…
View article: Investigating the Impact of Randomness on Reproducibility in Computer Vision: A Study on Applications in Civil Engineering and Medicine
Investigating the Impact of Randomness on Reproducibility in Computer Vision: A Study on Applications in Civil Engineering and Medicine Open
Reproducibility is essential for scientific research. However, in computer vision, achieving consistent results is challenging due to various factors. One influential, yet often unrecognized, factor is CUDA-induced randomness. Despite CUDA…
View article: Out of spuriousity: Improving robustness to spurious correlations without group annotations
Out of spuriousity: Improving robustness to spurious correlations without group annotations Open
Machine learning models are known to learn spurious correlations, i.e., features having strong relations with class labels but no causal relation. Relying on those correlations leads to poor performance in the data groups without these cor…
View article: Patch-based Intuitive Multimodal Prototypes Network (PIMPNet) for Alzheimer's Disease classification
Patch-based Intuitive Multimodal Prototypes Network (PIMPNet) for Alzheimer's Disease classification Open
Volumetric neuroimaging examinations like structural Magnetic Resonance Imaging (sMRI) are routinely applied to support the clinical diagnosis of dementia like Alzheimer's Disease (AD). Neuroradiologists examine 3D sMRI to detect and monit…
View article: Has this Fact been Edited? Detecting Knowledge Edits in Language Models
Has this Fact been Edited? Detecting Knowledge Edits in Language Models Open
Knowledge editing methods (KEs) can update language models' obsolete or inaccurate knowledge learned from pre-training. However, KEs can be used for malicious applications, e.g., inserting misinformation and toxic content. Knowing whether …
View article: CEval: A Benchmark for Evaluating Counterfactual Text Generation
CEval: A Benchmark for Evaluating Counterfactual Text Generation Open
Counterfactual text generation aims to minimally change a text, such that it is classified differently. Judging advancements in method development for counterfactual text generation is hindered by a non-uniform usage of data sets and metri…
View article: LLMs for Generating and Evaluating Counterfactuals: A Comprehensive Study
LLMs for Generating and Evaluating Counterfactuals: A Comprehensive Study Open
As NLP models become more complex, understanding their decisions becomes more crucial. Counterfactuals (CFs), where minimal changes to inputs flip a model's prediction, offer a way to explain these models. While Large Language Models (LLMs…
View article: Comprehensive Study on German Language Models for Clinical and Biomedical Text Understanding
Comprehensive Study on German Language Models for Clinical and Biomedical Text Understanding Open
Recent advances in natural language processing (NLP) can be largely attributed to the advent of pre-trained language models such as BERT and RoBERTa. While these models demonstrate remarkable performance on general datasets, they can strug…
View article: Corpus Considerations for Annotator Modeling and Scaling
Corpus Considerations for Annotator Modeling and Scaling Open
Recent trends in natural language processing research and annotation tasks affirm a paradigm shift from the traditional reliance on a single ground truth to a focus on individual perspectives, particularly in subjective tasks. In scenarios…
View article: Prototype-based Interpretable Breast Cancer Prediction Models: Analysis and Challenges
Prototype-based Interpretable Breast Cancer Prediction Models: Analysis and Challenges Open
Deep learning models have achieved high performance in medical applications, however, their adoption in clinical practice is hindered due to their black-box nature. Self-explainable models, like prototype-based models, can be especially be…
View article: PIPNet3D: Interpretable Detection of Alzheimer in MRI Scans
PIPNet3D: Interpretable Detection of Alzheimer in MRI Scans Open
Information from neuroimaging examinations is increasingly used to support diagnoses of dementia, e.g., Alzheimer's disease. While current clinical practice is mainly based on visual inspection and feature engineering, Deep Learning approa…
View article: A Second Look on BASS -- Boosting Abstractive Summarization with Unified Semantic Graphs -- A Replication Study
A Second Look on BASS -- Boosting Abstractive Summarization with Unified Semantic Graphs -- A Replication Study Open
We present a detailed replication study of the BASS framework, an abstractive summarization system based on the notion of Unified Semantic Graphs. Our investigation includes challenges in replicating key components and an ablation study to…
View article: The Queen of England is not England's Queen: On the Lack of Factual Coherency in PLMs
The Queen of England is not England's Queen: On the Lack of Factual Coherency in PLMs Open
Factual knowledge encoded in Pre-trained Language Models (PLMs) enriches their representations and justifies their use as knowledge bases. Previous work has focused on probing PLMs for factual knowledge by measuring how often they can corr…
View article: InfoLossQA: Characterizing and Recovering Information Loss in Text Simplification
InfoLossQA: Characterizing and Recovering Information Loss in Text Simplification Open
Text simplification aims to make technical texts more accessible to laypeople but often results in deletion of information and vagueness. This work proposes InfoLossQA, a framework to characterize and recover simplification-induced informa…