One Model, Many Behaviors: Training-Induced Effects on Out-of-Distribution Detection
Not on the map. Why?
Abstract (arXiv)
Out-of-distribution (OOD) detection is crucial for deploying robust and reliable machine-learning systems in open-world settings. Despite steady advances in OOD detectors, their interplay with modern training pipelines that maximize in-distribution (ID) accuracy and generalization remains under-explored. We investigate this link through a comprehensive empirical study. Fixing the architecture to the widely adopted ResNet-50, we benchmark 21 post-hoc, state-of-the-art OOD detection methods across 56 ImageNet-trained models obtained via diverse training strategies and evaluate them on eight OOD test sets. Contrary to the common assumption that higher ID accuracy implies better OOD detection performance, we uncover a non-monotonic relationship: OOD performance initially improves with accuracy but declines once advanced training recipes push accuracy beyond the baseline. Moreover, we observe a strong interdependence between training strategy, detector choice, and resulting OOD performance, indicating that no single method is universally optimal.
Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Builds on 69 of its 87 references that are in the map. First posted on arXiv 15 Jan 2026.
Builds on
69 of 87 references.
References · 69
- 2015Deep Residual Learning for Image Recognitioncited by 52,096
- 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 30,186
- 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,305
- 2021Masked Autoencoders Are Scalable Vision Learnerscited by 7,403
- 2021Emerging Properties in Self-Supervised Vision Transformerscited by 6,968
- 2017Towards Deep Learning Models Resistant to Adversarial Attackscited by 6,415
- 2015Rethinking the Inception Architecture for Computer Visioncited by 5,475
- 2017mixup: Beyond Empirical Risk Minimizationcited by 5,183
- 2017On Calibration of Modern Neural Networkscited by 4,960
- 2020Training data-efficient image transformers & distillation through attentioncited by 4,586
- 2017Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithmscited by 4,349
- 2019CutMix: Regularization Strategy to Train Strong Classifiers with Localizable Featurescited by 2,884
38 more
- 2020Unsupervised Learning of Visual Features by Contrasting Cluster Assignmentscited by 2,857
- 2020Supervised Contrastive Learningcited by 2,829
- 2016A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networkscited by 2,385
- 2019RandAugment: Practical automated data augmentation with a reduced search spacecited by 2,269
- 2013Describing Textures in the Wildcited by 2,125
- 2019Invariant Risk Minimizationcited by 1,854
- 2020The Many Faces of Robustness: A Critical Analysis of Out-of-Distribution Generalizationcited by 1,850
- 2021An Empirical Study of Training Self-Supervised Vision Transformerscited by 1,616
- 2018A Simple Unified Framework for Detecting Out-of-Distribution Samples and Adversarial Attackscited by 1,485
- 2018ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustnesscited by 1,432
- 2020Big Self-Supervised Models are Strong Semi-Supervised Learnerscited by 1,380
- 2017Enhancing The Reliability of Out-of-distribution Image Detection in Neural Networkscited by 1,328
- 2020Energy-based Out-of-distribution Detectioncited by 1,299
- 2021Generalized Out-of-Distribution Detection: A Surveycited by 802
- 2016Universal adversarial perturbationscited by 784
- 2014Deep Neural Networks are Easily Fooled: High Confidence Predictions for Unrecognizable Imagescited by 776
- 2022Out-of-Distribution Detection with Deep Nearest Neighborscited by 600
- 2021How to train your ViT? Data, Augmentation, and Regularization in Vision Transformerscited by 569
- 2021ReAct: Out-of-distribution Detection With Rectified Activationscited by 486
- 2019Scaling Out-of-Distribution Detection for Real-World Settingscited by 465
- 2018Bag of Tricks for Image Classification with Convolutional Neural Networkscited by 431
- 2020Generalized ODIN: Detecting Out-of-distribution Image without Learning from Out-of-distribution Datacited by 408
- 2022ViM: Out-Of-Distribution with Virtual-logit Matchingcited by 405
- 2021ResNet strikes back: An improved training procedure in timmcited by 368
- 2021Open-Set Recognition: a Good Closed-Set Classifier is All You Need?cited by 353
- 2021On the Importance of Gradients for Detecting Distributional Shifts in the Wildcited by 339
- 2018Why ReLU networks yield high-confidence predictions far away from the training data and how to mitigate the problemcited by 323
- 2020Do Adversarially Robust ImageNet Models Transfer Better?cited by 316
- 2022VOS: Learning What You Don't Know by Virtual Outlier Synthesiscited by 300
- 2022OpenOOD: Benchmarking Generalized Out-of-Distribution Detectioncited by 268
- 2021When Vision Transformers Outperform ResNets without Pre-training or Strong Data Augmentationscited by 260
- 2019Fast AutoAugmentcited by 243
- 2022Extremely Simple Activation Shaping for Out-of-Distribution Detectioncited by 212
- 2021A Simple Fix to Mahalanobis Distance for Improving Near-OOD Detectioncited by 207
- 2023OpenOOD v1.5: Enhanced Benchmark for Out-of-Distribution Detectioncited by 179
- 2021DICE: Leveraging Sparsification for Out-of-Distribution Detectioncited by 166
- 2023In or Out? Fixing ImageNet Out-of-Distribution Detection Evaluationcited by 134
- 2021PixMix: Dreamlike Pictures Comprehensively Improve Safety Measurescited by 131
50 most cited of 69 All 69 (JSON)
Built on by
Citing papers per year: 2026: 1.
1 paper.