Maximilian Ilse

Papers

Total Citations

H-Index

About

Maximilian Ilse is a researcher at the intersection of machine learning, causality, and domain generalization, with a focus on making models robust to distribution shifts. His most-cited work, "Selecting Data Augmentation for Simulating Interventions" (2020, 22 citations), tackles a fundamental challenge: models trained on observational data often fail in new environments due to spurious correlations. Ilse proposes a principled framework that uses data augmentation to simulate interventions, effectively decoupling domain-specific features from causal ones. This contribution bridges causal inference and practical deep learning, offering a systematic way to improve generalization without requiring interventional data. Beyond this, Ilse’s broader research explores how to leverage causal reasoning to design more reliable AI systems, particularly in high-stakes settings like medical imaging. His work is notable for its clarity in connecting theoretical causality to actionable algorithms, earning recognition among researchers working on out-of-distribution robustness. With a growing citation footprint, Ilse is helping shape a new generation of models that learn not just correlations, but the underlying mechanisms driving data—a critical step toward trustworthy artificial intelligence.