Workshop Program

Time Session
09:00 - 09:30 People join & Welcome
09:30 - 10:00 Invited Talk by Jes Frellsen
An Isotropic Approach to Efficient Uncertainty Quantification with Gradient Norms
10:00 - 10:30 Invited Talk by Teodora Gagaleska
Diagnosing Multimodal LLMs: Contradictions, Modality Bias, and What Benchmarks Miss
10:30 - 11:00 Coffee Break
11:00 - 11:30 Invited Talk by Christian Hardmeier
Towards Anthropomimetic Uncertainty: Desiderata for Uncertainty Communication in Large Language Models
11:30 - 11:45 Contributed Oral Presentations (Accepted Full Papers)
2 × 7 min talks + Q&A
  • Does Marginal Coverage Guarantee Class-Conditional Safety for Zero-Shot VLMs Under Shift?
  • Uncertainty-Aware Trajectory Forecasting from Imperfect Tracking
11:45 - 12:15 Contributed Spotlights (Accepted Extended Abstracts)
7 × 3 min presentations
  • Modality-Aware Out-of-Distribution Detection for Multi-Modal Action Recognition
  • PUF: Plug-and-Play Uncertainty-Aware Fusion for Online 3D Scene Graph Generation
  • Auto3R: Automated 3D Reconstruction and Scanning via Data-driven Uncertainty Quantification
  • Open-Set Domain Adaptation Under Background Distribution Shift
  • Decompose, Compare, and Decide: Multimodal LLMs are Implicit Few-Shot Learners
  • Ensembling Pruned Attention Heads For Uncertainty-Aware Efficient Transformers
  • Rank-Dropout LoRA for Parameter-Efficient Uncertainty Estimation
12:30 - 13:15 Lunch
13:15 - 14:30 Poster Session
14:30 - 15:00 Invited Talk by Yarin Gal
TBA
15:00 - 15:30 Coffee Break
15:30 - 16:00 Invited Talk by Noa Garcia
Invisible bias: efffects on evaluation
16:00 - 16:30 Invited Talk by Matthieu Cord
Analyzing Internal Representations of Multimodal LLMs
16:30 - 17:00 Invited Talk by Lorenzo Baraldi
Looking Inside Multimodal Foundation Models: Representations as the Interface for Control and Reliability