Inspiration Exchange #54

We will feature the following papers:
Hyperparameter Trajectory Inference with Conditional Lagrangian Optimal Transport
     Harry Amad, Mihaela van der Schaar
Eliciting Numerical Predictive Distributions of LLMs Without Auto-Regression
Julianna Piskorz, Kasia Kobalczyk, Mihaela van der Schaar
Deep Hierarchical Learning with Nested Subspace Networks
Paulius Rauba, Mihaela van der Schaar
Fact-Augmented Lookahead for LLM Agents: Simple Online Memory, No FinetuningSamuel Holt, Max Ruiz Luyten, Thomas Pouplin, Mihaela van der Schaar
These papers represent a subset of our accepted papers at ICLR and AISTATS 2026. Several of the other accepted papers have already been presented and discussed in previous Inspiration Exchange sessions. A summary graphic of all accepted papers is included below and the links to our website posts are here: https://www.vanderschaar-lab.com/aistats-2026/ https://www.vanderschaar-lab.com/iclr-2026/

The presentations will be followed by a panel discussion with industry experts. As always, the session will be interactive, and we strongly encourage discussion and questions from participants.