Pith. sign in

REVIEW 9 cited by

Scaling Out-of-Distribution Detection for Real-World Settings

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 1911.11132 v4 pith:BZAISUEE submitted 2019-11-25 cs.CV cs.LG

Scaling Out-of-Distribution Detection for Real-World Settings

Dan Hendrycks , Steven Basart , Mantas Mazeika , Andy Zou , Joe Kwon , Mohammadreza Mostajabi , Jacob Steinhardt , Dawn Song This is my paper
classification cs.CV cs.LG
keywords settingsout-of-distributionanomalydetectionlarge-scalesegmentationsimplebenchmark
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
read the original abstract

Detecting out-of-distribution examples is important for safety-critical machine learning applications such as detecting novel biological phenomena and self-driving cars. However, existing research mainly focuses on simple small-scale settings. To set the stage for more realistic out-of-distribution detection, we depart from small-scale settings and explore large-scale multiclass and multi-label settings with high-resolution images and thousands of classes. To make future work in real-world settings possible, we create new benchmarks for three large-scale settings. To test ImageNet multiclass anomaly detectors, we introduce the Species dataset containing over 700,000 images and over a thousand anomalous species. We leverage ImageNet-21K to evaluate PASCAL VOC and COCO multilabel anomaly detectors. Third, we introduce a new benchmark for anomaly segmentation by introducing a segmentation benchmark with road anomalies. We conduct extensive experiments in these more realistic settings for out-of-distribution detection and find that a surprisingly simple detector based on the maximum logit outperforms prior methods in all the large-scale multi-class, multi-label, and segmentation tasks, establishing a simple new baseline for future work.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 9 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. When Confidence Lacks Concepts: Interpretable OOD Detection via Representation Perturbations

    cs.LG 2026-06 unverdicted novelty 6.0

    Proposes using sparse autoencoders to extract class-conditioned concept vectors, then measuring logit stability under targeted perturbations as an interpretable OOD signal for deep networks in medical imaging.

  2. Debiased Negative Mining Improves Out-of-distribution Detection with Pre-trained Vision-Language Models

    cs.LG 2026-05 unverdicted novelty 6.0

    Debiased negative mining via Monte-Carlo sampling from ID labels and unlabeled wild data improves OOD detection with VLMs and achieves new state-of-the-art results.

  3. Open Set Face Forgery Detection via Dual-Level Evidence Collection

    cs.CV 2025-12 unverdicted novelty 6.0

    DLED reformulates open-set face forgery detection as an uncertainty estimation task and uses dual-level spatial-frequency evidence collection to identify novel fake categories, claiming 20% average gains over baselines.

  4. ConjNorm: Tractable Density Estimation for Out-of-Distribution Detection

    cs.LG 2024-02 unverdicted novelty 6.0

    ConjNorm reframes OOD detection score design as optimizing norm p in an exponential family density model via a Bregman divergence theorem, with a tractable Monte Carlo estimator, claiming SOTA gains on CIFAR-100 and I...

  5. Face-trace: Open-Set Attribution and Progressive Discovery of Synthetic Face Generators

    cs.CV 2026-07 conditional novelty 5.0

    A pipeline combining frozen I-JEPA embeddings, energy-based OOD rejection, and HDBSCAN clustering achieves 87.74% purity in discovering unknown synthetic face generators, extended to an incremental setting with 99.23%...

  6. Road-Aware Anomaly Segmentation with Query-Guided Polygons and CLIP in Autonomous Driving

    cs.CV 2026-07 conditional novelty 5.0

    Road-query polygons plus CLIP zero-shot filtering turn Mask2Former soft masks into stronger training-free road anomaly segmentations, topping Maskomaly AP on Fishyscapes LostAndFound.

  7. Learning Hyperspherical Time-Frequency Representations for Time-Series Out-of-Distribution Detection

    cs.LG 2026-05 unverdicted novelty 5.0

    Hyperspherical time-frequency representations learned via von Mises-Fisher likelihood improve OOD detection on UCR and UEA archives using k-NN and Mahalanobis scores over contrastive baselines.

  8. VOLTA: The Surprising Ineffectiveness of Auxiliary Losses for Calibrated Deep Learning

    cs.LG 2026-04 unverdicted novelty 5.0

    VOLTA, consisting of a deep encoder with learnable prototypes plus cross-entropy and post-hoc temperature scaling, matches or exceeds ten UQ baselines in accuracy, achieves lower expected calibration error, and perfor...

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

    cs.LG 2025-11 unverdicted novelty 5.0

    Benchmark across architectures and shift regimes finds OOD detector rankings shift with representation collapse; proposes NC-based shortlist predictor and PCA filter without extra OOD data.