Visual Grounding First, Multimodal In-context Learning Follows

Minhyuk Seo, Minjae Lee, Chaeeun Lee, Wei Lin and 3 more

Sydney Poster Session 2, Tue, Dec 8, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026

$\mathcal{P}$Torch: Narrowing the Gap Between Projection and Gradient-Based Learning

Jeff Cyuzuzo Jambé, Jan Quan, Panagiotis Patrinos

Sydney Poster Session 4, Wed, Dec 9, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026

Dual-Space Preconditioning for Variational Inequalities and Root-Finding Problems

Jan Quan, Konstantinos Oikonomidis, Alexander Bodard, Panagiotis Patrinos

Sydney Poster Session 4, Wed, Dec 9, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026

When does the noise schedule matter? A spectral classification of diffusion training objectives

Ali Raza

Sydney Poster Session 1, Tue, Dec 8, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026

Approximate Bayesian inference with exchangeable distributions for neurosymbolic AI

Lennert De Smet

Paris Poster Session 1, Wed, Dec 9, 12:30 PM–2:30 PM, Paris Poster Hall · Published 2026

Online Decision-Focused Learning under Semi-Bandit Feedback

Aabhash Dhakal, Tim Lachner, Jayanta Mandi, Marco Foschini and 1 more

Sydney Poster Session 2, Tue, Dec 8, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026

PhyTS: A Benchmark for Scientific Time Series

Benedict Armstrong, Jeroen Audenaert, Hannah P Binney, Alice Cheng and 22 more

Sydney Poster Session 6, Thu, Dec 10, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026

Constrained Stochastic Spectral Preconditioning Converges for Nonconvex Objectives

Proximal preconditioned stochastic gradient methods extend Muon/Scion to nonconvex constrained optimization with heavy-tailed noise convergence and faster variance-reduced variants.

Konstantinos Oikonomidis, Jan Quan, Kimon Antonakopoulos, Antonio Silveti-Falls and 2 more

Sydney Poster Session 3, Wed, Dec 9, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026

EM-NeSy: Expectation Maximization for Neurosymbolic Learning

EM-NeSy casts neurosymbolic learning as expectation-maximization to enable approximate symbolic reasoning without requiring differentiable reasoning components.

Annegret Seibt, Luc De Raedt, Giuseppe Marra

Paris Poster Session 1, Wed, Dec 9, 12:30 PM–2:30 PM, Paris Poster Hall · Published 2026

On the Convergence of Multicalibration Gradient Boosting

Multicalibration gradient boosting converges at O(1/sqrt(T)) with linear rates under smoothness, plus adaptive guarantees backed by experiments.

Daniel Haimovich, Fridolin Linder, Lorenzo Perini, Niek Tax and 1 more

Sydney Poster Session 1, Tue, Dec 8, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026

NeuroAtlas: Benchmarking Foundation Models for Clinical EEG and Brain-Computer Interfaces

NeuroAtlas benchmarks EEG foundation models across 42 datasets and finds they largely match generic time-series models without delivering unified clinical EEG performance.

Konstantinos Kontras, Trui Osselaer, Stylianos G Mouslech, Angeliki I. Karaiskou and 11 more

Sydney Poster Session 5, Thu, Dec 10, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026

Divide et Calibra: Multiclass Local Calibration via Vector Quantization

Divide et Calibra uses vector quantization to learn shared, region-specific multiclass calibration maps that improve local calibration without reducing latent dimensions.

Cesare Barbera, Lorenzo Perini, Giovanni De Toni, Andrea Passerini and 1 more

Paris Poster Session 6, Fri, Dec 11, 2:30 PM–4:30 PM, Paris Poster Hall · Published 2026

EchoPrune: Interpreting Redundancy as Temporal Echoes for Efficient VideoLLMs

EchoPrune treats redundant video tokens as temporal echoes and prunes them via query relevance and reconstruction error, letting VideoLLMs process up to 20x more frames for +8.6% accuracy and 5.6x faster prefilling.

Jiameng Li, Minye Wu, Jiezhang Cao, Aleksei Tiulpin and 1 more

Sydney Poster Session 5, Thu, Dec 10, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026

Querying Counterfactuals on Tissue Graphs with Supervised Disentanglement

Cellina defines tissue graph counterfactuals as spatial edge or node interventions and uses supervised disentanglement to separate intrinsic cell states from context, outperforming competitors across millions of cells and revealing cancer subdomains.

Abdul Moeed, Stefan Schrod, Martin Rohbeck, Marc J Bonder and 3 more

Sydney Poster Session 3, Wed, Dec 9, 10:00 AM–1:00 PM, Hall 1-4 · Published 2026

CORTEG: Foundation Models Enable Cross-Modality Representation Transfer from Scalp to Intracranial Brain Recordings

CORTEG adapts pretrained scalp-EEG foundation models to intracranial ECoG via cross-modality transfer, enabling rapid patient calibration with competitive or superior decoding performance.

Liuyin Yang, Qiang Sun, Bob Van Dyck, Eva C Merino and 1 more

Sydney Poster Session 6, Thu, Dec 10, 5:00 PM–8:00 PM, Hall 1-4 · Published 2026

Neurosymbolic Object-Centric Learning with Distant Supervision

DeepObjectLog integrates object-centric encoding with probabilistic logic to learn object-level arguments from global labels, achieving stronger out-of-distribution generalization on visual reasoning tasks.

Stefano Colamonaco, David Debot, Giuseppe Marra

Paris Poster Session 3, Thu, Dec 10, 12:30 PM–2:30 PM, Paris Poster Hall · Published 2026