Projects from Kai Biermeier, M.Sc.
TRR 318-2 - Project A05: Contextualized and online parametrization of scaffolding in human–robot explanatory dialog
Negation and verbal contrast are important means of scaffolding. When negations are provided during explaining joint actions, they link the present knowledge about actions to those that have been performed in the past. We continue to develop scaffolding strategies for a dialog with a social robot in a joint action setting. Our long-term objective ...
Duration: 01/2026 - 06/2029
TRR 318-2 - Project Z: Central tasks of the Collaborative Research Center
Z is the central management hub of TRR 318: the project coordinates the budget, ensures internal communication, and organizes central events. In addition, Z pursues several missions aimed at best possible conditions, processes, and outcomes for the TRR’s research: offering an attractive interdisciplinary environment with empirical research support, ...
Duration: 01/2026 - 06/2029
TRR 318-2 - Project RTG: Integrated Research Training Group
The RTG structures the further training of early career researchers and prepares them for their interdisciplinary research work by familiarizing them with different scientific practices, disciplinary cultures, methods, attitudes and values and by fostering communication across disciplines or discourse communities including the public. The RTG aims ...
Duration: 01/2026 - 06/2029
TRR 318-2 - Project WIKO: Questions about explainable technology
WIKO integrates AI explanation research with public outreach, now focusing on advancing eXplainable AI (XAI) Literacy to foster understanding and reflection on transparent AI. Through self conceptualized and conducted Co-Construction workshops (CCWS), participants from different public sectors actively engage with AI systems, promoting critical ...
Duration: 01/2026 - 06/2029
TRR 318-2 - Project INF: Retrieval-augmented information provision
INF provides foundations for all projects. These include services and infrastructure for data access, evaluation, and management as well as cross-cutting research questions. In the first funding period, the focus was on assessing the quality of explanations, modeling explanatory dialog sequences as well as ontology and tool development. In the ...
Duration: 01/2026 - 06/2029
TRR 318-2 - Project C07: Co-construction-following large language models for explaining
C07 studies the adjustment of LLM training processes and knowledge integration to enable the successful co-construction of explanations. Scaffolding and monitoring are operationalized via a novel fine-tuning process that teaches LLMs to explain dialogically while modeling dialog states. A novel co-construction-aware RAG approach exploits these ...
Duration: 01/2026 - 06/2029
TRR 318-2 - Project C05: Creating explanations in collaborative human–machine knowledge exploration
C05 investigates how humans and AI systems can jointly explore a decision space to make an explainable and well-founded decision. We focus on diagnostic decision-making in medicine and combine empirical studies on physicians' exploration strategies and cognitive biases, as well as on assessments of ex-planations and argument strength, with the ...
Duration: 01/2026 - 06/2029
TRR 318-2 - Project C04: Metaphors as an explanation tool
Explanations of abstract explananda (such as AI systems) need metaphors to relate new, abstract knowledge to more concrete concepts. In the first phase, we studied how metaphors highlight and hide aspects to foster understanding, how to model their agency structure, and how they are used for explaining. In the second, we will investigate how to ...
Duration: 01/2026 - 06/2029
TRR 318-2 - Project C03: Interpretable machine learning—Explaining change
C03 develops methods to explain how and why machine learning models adapt over time. Our project makes AI systems more transparent in dynamic settings by designing efficient, expressive explanations. This includes novel approaches for real-time explanation and higher-order Shapley interactions. In the second funding period, we aim for ex-ante ...
Duration: 01/2026 - 06/2029
TRR 318-2 - Project C02: Interactive learning of explainable, situation-adapted decision models
C02 aims to optimize ML-based decision support systems and human-AI interaction in order to improve decision-making. We focus on explainable and transparent models that take the decision context into account. We are focusing on prescriptive decision-making with partial feedback, where we not only consider the long-term impact of decisions but also ...
Duration: 01/2026 - 06/2029