mlsanityData to 7 Oct 2026

Autoencoders for Anomaly Detection are Unreliable

arXiv 2501.13864 ↗First posted 23 Jan 2025cs.LG cs.AIcited by 9

Not on the map. Why?

Abstract (arXiv)

Autoencoders are frequently used for anomaly detection, both in the unsupervised and semi-supervised settings. They rely on the assumption that when trained using the reconstruction loss, they will be able to reconstruct normal data more accurately than anomalous data. Some recent works have posited that this assumption may not always hold, but little has been done to study the validity of the assumption in theory. In this work we show that this assumption indeed does not hold, and illustrate that anomalies, lying far away from normal data, can be perfectly reconstructed in practice. We revisit the theory of failure of linear autoencoders for anomaly detection by showing how they can perfectly reconstruct out of bounds, or extrapolate undesirably, and note how this can be dangerous in safety critical applications. We connect this to non-linear autoencoders through experiments on both tabular data and real-world image data, the two primary application areas of autoencoders for anomaly detection.

Key facts. Cited by 9 papers in the map (data through 7 Oct 2026). Builds on 16 of its 37 references that are in the map. First posted on arXiv 23 Jan 2025.

Builds on

16 of 37 references.

References · 16

  1. 2017Towards Deep Learning Models Resistant to Adversarial Attackscited by 6,326
  2. 2018Real-world Anomaly Detection in Surveillance Videoscited by 570
  3. 2019Memorizing Normality to Detect Anomaly: Memory-augmented Deep Autoencoder for Unsupervised Anomaly Detectioncited by 560
  4. 2018Do Deep Generative Models Know What They Don't Know?cited by 443
  5. 2022A Unified Model for Multi-class Anomaly Detectioncited by 267
  6. 2018Improving Reconstruction Autoencoder Out-of-distribution Detection with Mahalanobis Distancecited by 74
  7. 2023Hierarchical Vector Quantized Transformer for Multi-class Unsupervised Anomaly Detectioncited by 66
  8. 2022Rethinking Reconstruction Autoencoder-Based Out-of-Distribution Detectioncited by 59
  9. 2021Hierarchical VAEs Know What They Don't Knowcited by 58
  10. 2019Robust Anomaly Detection in Images using Adversarial Autoencoderscited by 38
  11. 2021Learning Not to Reconstruct Anomaliescited by 31
  12. 2020ARAE: Adversarially Robust Training of Autoencoders Improves Novelty Detectioncited by 27
4 more
  1. 2021Autoencoding Under Normalization Constraintscited by 26
  2. 2024Rethinking Autoencoders for Medical Anomaly Detection from A Theoretical Perspectivecited by 22
  3. 2023Unsupervised anomaly detection algorithms on real-world data: how many do we need?cited by 20
  4. 2024Exploiting Autoencoder's Weakness to Generate Pseudo Anomaliescited by 4

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 4; 2026: 5.

9 papers.

  1. 2026Failure Identification in Imitation Learning Via Statistical and Semantic Filteringcited by 6
  2. 2026Toward Faithful Explanations in Acoustic Anomaly Detectioncited by 1
  3. 2025Risk-Based Thresholding for Reliable Anomaly Detection in Concentrated Solar Power Plantscited by 0
  4. 2025Applying Deep Learning to Anomaly Detection of Russian Satellite Activity for Indications Prior to Military Activitycited by 0
  5. 2025Enhancing ECG Classification Robustness with Lightweight Unsupervised Anomaly Detection Filterscited by 0
  6. 2025Hybrid Quantum-Classical Autoencoders for Unsupervised Network Intrusion Detectioncited by 0
  7. 2026Modeling Normal Is All You Need: Joint Latent Clustering for Anomaly Detection in Multimodal Cyber-Physical Systemscited by 0
  8. 2026Cross-Domain Industrial Fault Detection by Causal Mechanism Monitoringcited by 0
  9. 2026When Does Unsupervised Learning Succeed or Fail? A PoS Perspective on Reconstruction-Based Anomaly Detectioncited by 0