exCV Workshop at ECCV 26 · 2026

We need theory-grounded explainability for abstract tasks to quantify model bias

Katharina Prasse, Margret Keuper

Explainable AI Model Bias Theory-Grounded Evaluation

Why this publication matters

Explaining a prediction about an abstract idea requires more than highlighting a patch of an image. This position argues that explanations should connect to well-defined, context-specific concepts. That connection matters if researchers want to measure model bias rather than merely produce explanations that look plausible.

Abstract

The original abstract will be added when the full paper is available.

Figures

Additional figures and captions will be added when the full paper is available.

Cite this paper

Download .bib

@inproceedings{prasse2026weneedtheory95,
  title = {{We need theory-grounded explainability for abstract tasks to quantify model bias}},
  author = {Katharina Prasse and Margret Keuper},
  booktitle = {exCV Workshop at ECCV 26},
  year = {2026},
  url = {https://openreview.net/forum?id=Oz2id4l0SX}
}

Figures and abstract are reproduced from the linked research sources. Credit remains with the authors and publishers.