Felix Gröner

Publications

Highlighted Technical Projects

Select one to view details. Hover for color images.

Loading…

Other Unpublished Projects

• People and LLMs Learning Each Other's Competencies

  Can people accurately calibrate their expectations of LLMs in various domains? Can LLMs identify whether they or the individual prompting them is more competent?

  Expectations Safety Mental Models

• LLMs Inferring Individual Preferences From Edits to Their Outputs

  Do LLM-generated summaries improve if preferences are inferred from people's revisions of previous summaries?

  Personalization Methodology

• Prompting GPT-4 for Intelligence Analysis Reporting Standards

  Can GPT-4 adhere to the intelligence community's best practices for reporting and does that help analysts to assess the reliability of the system?

  Evaluation Prompt Engineering Intelligence Analysis

• Designing LLM UIs for Appropriate Reliance

  What user interface elements (e.g., warnings, confidence bars, cognitive-forcing functions) can assist people in correctly calibrating their reliance on LLMs?

  Design Appropriate Reliance Dissertation

• AI Images for Cognitive Psychology Research

  What computational metrics explain people's internal representations of abstract art? And can we use synthetic AI-generated mixes of original artworks? Check out this interactive visualization .

  Image Generators Cognitive Psychology

• What Factors Influence People's Perceptions of AI Recommendations?

  We showed global (linear surrogate) and local (LIME) explanations for an AI system's recommendations. This distinction turned out to not have as large of an effect as the participant's self-perceived expertise.

  Classifiers Explainable AI (XAI)

Professional Experience