Open Topics

Ongoing

Radhika Yadav (with Sherzod Hakimov) – MSc thesis @ Uni Potsdam: Measuring and Mitigating Overreliance on Explanations by Fact-Checking Systems (Title TBD)

Yi-Sheng Hsu (with Manfred Stede) – MSc thesis @ Uni Potsdam: Evaluating LLM-generated Explanatory Utterances through Dialogue Completion

Completed

Yi-Sheng Hsu (with Sherzod Hakimov) – MSc project @ Uni Potsdam: Engineering LLM-generated Explanations with Metric-based Readability Control

Maximilian Bleick (with Aljoscha Burchardt) – BA thesis @ TU Berlin: An Investigation of LLM Chatbots Concerning the Echo Chamber Effect

Qianli Wang (with Leonhard Hennig) – MSC thesis @ TU Berlin: A Singular LLM Is All You Need for Dialogue-based Explanation Regarding NLP Tasks

Konstantin Biskupski (with Eleftherios Avramidis) – MSc thesis @ TU Berlin: Quality estimation of machine-translated texts with fine-grained classification of errors

Kiran Rohra (with Philippe Thomas) – MSc thesis @ TU Berlin: Comparative error analysis of biomedical image labelling and captioning models

Ajay Madhavan Ravichandran (with Philippe Thomas) – MSc thesis @ TU Berlin: Evaluating text quality of generated radiology reports

Mika Rebensburg (with Tim Polzehl & Stefan Hillmann) - BSc thesis @ TU Berlin : Automatic Evaluation of Chatbot Dialogs Using Pre-Trained Language Models in the Customer Support Domain

Daniel Fernau (with Tim Polzehl & Stefan Hillmann) – MSc thesis @ TU Berlin: Towards Adaptive Conversational Agents: Fine-tuning Language-Models for User Classification to enhance Usability

Research assistant supervisions

Ongoing

Yi-Sheng Hsu – Explanation Dialogues; Rationalization with Metric-based Readability Control

Maximilian Dustin Nasert – Data-based Interpretability

Completed

Christopher Ebert – Saliency map verbalization and data-based interpretability

Qianli Wang – Interactive NLP model exploration through dialogue systems

João Lucas Mendes de Lemos Lins – Instructional explanations

Sahil Chopra – Rationale generation for dialogue-based explanations

Ajay Madhavan Ravichandran – Conceptualizing dialogue-based explanations