Optimal Transport Reweighting for Robust Learning under Spurious Correlations and Label Noise

POTER uses optimal transport geometry between training and reference distributions to downweight mislabeled or shortcut-aligned samples, achieving state-of-the-art worst-group accuracy with a single training stage.

Sung Ho Jo, Seonghwi Kim, Wonsang Yun, Minwoo Chae

Sydney Poster Session 2, Tue, Dec 8, 5:00 PM–8:00 PM, Hall 1-4 · Published Oct 1, 2026 · 0 citations

Dataset Identity, Not Novelty: The Source of an Inflated OOD Detection Gain

Post-hoc OOD detectors inflate reported gains by fitting dataset identity rather than novelty, leaving only a single constant fit valid.

Donghoon Lee, Shinjin Kang

Published Oct 1, 2026 · 0 citations

When to Trust a PFN: Detecting Harmful Shift in Tabular Foundation Models

Viet Nguyen, Herman Bergström, Stephan Rabanser, Rahul Krishnan

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

Robust-by-Design Distributional Learning from Contaminated Samples

Nevena Gligić, Arya Farahi

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

Decoupling Label Shift and Surrogate Gradient Errors for Robust Federated Spiking Neural Networks

Zouquan Chen, Yifei Yang, Lidong Zheng, Zhiming Fang and 1 more

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

Robust Satisficing Ensemble: Scalable Model Aggregation Under Distribution Shifts

Ahmet Faruk Cetinkaya, Enes Ağırman, Cem Tekin

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

DEDCA: Test-Time Adaptation for Generalized AI-Generated Image Detection

Hao Tan, Qimin Zhang, Lihuang Fang, Kebing Jin and 1 more

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

D-GAP: Improving Out-of-Domain Robustness via Dataset-Agnostic and Gradient-Guided Augmentation in Amplitude and Pixel Spaces

Ruoqi Wang, Haitao Wang, Shaojie Guo, Qiong Luo

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

Adaptive Covariance and Multi-Layer Alignment for Out-of-Distribution Detection

Haoyang Su, Qi Chen, Max Gutbrod, Johan Verjans and 1 more

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

Strengthen Out-of-Distribution Detection via Adaptive Mahalanobis Gap

Kunpeng Sui, Rundong He, Jie Su

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

Orthogonal Origin Parking: Decoupling Lorentz Manifolds for Robust OOD Generalization

Peter J Kampen, Anders N Christensen, Morten Rieger Hannemose, Anders Dahl and 1 more

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

Distributionally Robust Mixture-of-Experts Training

Xin Teng, Muxiao Li, Hongyi Wen

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

Improving Context-Shift Robustness of Convolutional Networks via Context-Regularized Cross-Entropy

Jinen Setpal, Chaoyue Liu

Atlanta Poster Session 5, Fri, Dec 11, 10:00 AM–1:00 PM, Hall C1 · Published 2026

REACT: A Lightweight Reliability-Aware Framework for Spatio-Temporal Out-of-Distribution Prediction

Yongfeng Su, Ziquan Fang, Jihua Yang, Wei Shao and 1 more

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

Out-of-Distribution Detection in Continual Learning

Nimeshika Udayangani Hewa Dehigahawattage, Sarah Erfani, Flora Salim, Christopher Leckie

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

Beyond the Training Distribution: Evaluating Predictions Under Distribution Shift and Selection Bias

A double machine learning procedure estimates black-box prediction risk under covariate shift and selective labels, tracking true target risk more accurately than single-source methods.

Annie Ulichney, Amanda L Coston

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

Perturb and Correct: Post-Hoc Ensembles using Affine Redundancy

Perturb-and-Correct builds diverse predictors via hidden-layer perturbation and affine correction, yielding strong in-distribution versus out-of-distribution tradeoffs from one pretrained model.

Eleanor Quint

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

Benchmarking and Improving Monitors for Out-Of-Distribution Alignment Failure in LLMs

MOOD benchmark shows guard models fail to detect out-of-distribution alignment failures, but combining them with Mahalanobis and perplexity detectors improves recall from 39% to 45% and scales positively.

Dylan Feng, Pragya Srivastava, Anca Dragan, Cassidy Laidlaw

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

LLM Alignment--Utility Asymmetry under Semantic-Preserving Transformations

Synthetic semantic-preserving transformations reveal alignment-utility asymmetry: LLMs retain task utility on shifted inputs but suffer sharp alignment failures, with harmful rates surging over 40 points despite minimal capability loss.

Mohan Li, Chengyu Yu, Francesco Sovrano, Marc Langheinrich and 1 more

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

A Systematic Analysis of Out-of-Distribution Detection Under Representation and Training Paradigm Shifts

A systematic benchmark shows out-of-distribution detector competitiveness depends mainly on learned representations rather than score design, with neural collapse metrics predicting top detector choices without extra out-of-distribution data.

Claudio César Claros-Olivares, Austin Brockmeier

Atlanta Poster Session 3, Thu, Dec 10, 10:00 AM–1:00 PM, Hall C1 · Published 2026