News
- Jul 2026Attending ICML 2026 in Seoul, South Korea.
- Jul 2026Attending the Oxford Machine Learning School.
- Jun 2026New preprint: OncoSynth, a causally-aware diffusion framework for synthetic patient data generation in oncology.
- Jun 2026Presented SurvDiff as a talk at the Helmholtz AI Conference (HAICON26), Munich.
- May 2026New preprint: ConfoundingSHAP, a Shapley-based method for quantifying confounding strength in causal inference.
- May 2026New preprint: Amortizing Causal Sensitivity Analysis via Prior Data-Fitted Networks.
- Apr 2026SurvDiff accepted at ICML 2026 as a spotlight paper, top 2.2% of all 24,000+ submissions.
- Sep 2025New preprint: SurvDiff, a diffusion model for synthetic data generation in survival analysis.
- Jun 2025Started my PhD at LMU Munich under the supervision of Stefan Feuerriegel.
Publications
First author
ConfoundingSHAP: Quantifying Confounding Strength in Causal Inference
Contributed
Causal Methods for LLM Development and Evaluation
Amortizing Causal Sensitivity Analysis via Prior Data-Fitted Networks
OncoSynth: Synthetic Data Generation for Treatment Effect Estimation in Oncology
Impressum · Marie Brockschmidt · Ludwigstr. 28 RG, 80539 München · marie [dot] brockschmidt [at] lmu [dot] de