Lukas Ellinger
Research Group Social Computing. Technical University Munich.
I am a first-year PhD candidate at the Research Group Social Computing at the Technical University of Munich. My academic background is in Informatics, with a focus on Natural Language Processing and Large Language Models. I am broadly interested in how people and AI interact through language, how meaning is negotiated, and how communication breaks down when intent is unclear or context is misunderstood.
My research investigates how Large Language Models (LLMs) interpret and adapt to conversational context, particularly when user intent is underspecified or ambiguous. I study how LLMs model and maintain common ground, both as participants in a conversation and as external observers interpreting interactions between others. My work focuses on improving these representations to enable more reliable inference of message intent, robust handling of ambiguity, and mitigation of framing effects.