Laura Kopf is a PhD candidate at Technische Universität Berlin (TU Berlin), pursuing a doctorate in Computer Science, and a research associate at BIFOLD. Her research explores the internal representations of neural networks, with a particular focus on understanding which concepts are encoded within these models. Laura works on developing automated interpretability methods and evaluating the reliability of such approaches.

Selected publications

• Capturing Polysemanticity with PRISM: A Multi-Concept Feature Description Framework [NeurIPS 2025]
• Interpreting Language Models Through Concept Descriptions: A Survey [BlackboxNLP Workshop @ EMNLP 2025]
• CoSy: Evaluating Textual Explanations of Neurons [NeurIPS 2024]
• Labeling Neural Representations with Inverse Recognition [NeurIPS 2023]