DAG learning with 2-nodes is sensitive to scaling. In #ICLR2024 we show it's true for arbitrary dim. variables & log likelihood or squared error losses can be minimal for the wrong DAG, even in non-linear settings: openreview.net/forum?id=gwbQ2… Jonas Seng Matej Zečević Kristian Kersting

News from our CSG Data Science and Machine Learning: Jonas Seng presents the training „Methods of AutoML and how to utilize them“: 📅 March 21, 2024, 9.00 am - 1.00 pm, online Register now: eveeno.com/methods_of_aut… #NHR4CES #HPC #ML #MachineLearning #AI #Service #Training

The first “ConCon” challenge, tackling continual confounding, has just been launched!💡 Submit 3-page reports (non-archival) as part of the third Cont. Causal. Bridge at AAAI@2025. (Early deadline is Dec 6, 2024 AoE) Check out the full announcement here: continualcausality.org/challenge/

Thrilled to share our #ICLR2025 work on Meta-Causal States! 🌟 Causal graphs evolve with dynamic systems & agent actions. We show how to cluster causal models by qualitative behavior, revealing hidden dynamics & emergent relationships 🚀 #Causality #ML arxiv.org/abs/2410.13054

I'm excited to present our spotlight on meta-causal models at #ICLR2025 next week. We model evolving causal graphs in dynamic systems. Applications to inference and attribution of agent actions. Paper: openreview.net/forum?id=J9Vog… Visit our poster #441 during the Sat 3pm session.

Horizontal, vertical, hybrid? Data heterogeneity is tough in federated learning! 🔥In "Scaling Probabilistic Circuits via Data Partitioning" - accepted UAI 2025 - we introduce a unified view through aggregation of learned client distributions: arxiv.org/abs/2503.08141

Although he could be surfing right now, Jonas Seng is busy presenting our work on the interplay between probabilistic circuits and federated learning UAI 2025 in Rio 🇧🇷. I am here till Friday and will be happy to chat about cauaslity, PCs or anything under the sun.