3 papers
cs.LG2026
Learning Transferable Policies from Action-free Time Series Through Dynamical Embeddings
Niklas Emonds, Georgia Koppe
Learning control from action-free recordings is challenging because intervention effects are unobserved and policies may exploit errors in reconstructed dynamics. We present a hier…
cs.LG2026
The Dynamic-Probabilistic Consistency Gap in Chaotic Surrogate Modeling
Andre Herz, Matthijs Pals, Daniel Durstewitz +1
Dynamical systems reconstruction (DSR) aims to learn surrogate models that capture the dynamics underlying time-series data. Reliably deploying these surrogates requires uncertaint…
cs.LG2026
Teacher Forcing as Generalized Bayes: Optimization Geometry Mismatch in Switching Surrogates for Chaotic Dynamics
Andre Herz, Daniel Durstewitz, Georgia Koppe
Identity teacher forcing (ITF) enables stable training of deterministic recurrent surrogates for chaotic dynamical systems and has been highly effective for dynamical systems recon…