The paper studies a class of tabular anomaly detectors that train a neural map toward a fixed target using squared‑error loss, noting that improved optimization can paradoxically worsen detection performance—a phenomenon termed convergence collapse. It demonstrates that this collapse arises because the learned map extrapolates beyond the training data, erasing residual signals on anomalies. To address this, the authors introduce Kernel Contraction Matching (KCM), a closed‑form, training‑free kernel‑based baseline, and propose the Kernel‑Anchored Regularizer (KAR), which penalizes deviations from a kernel‑weighted average of training targets, thereby mitigating collapse and improving AUROC during extended training.
By Jos\'e Lucas De Melo Costa, Fabrice Popineau, Arpad Rimmel, Bich-Li\^en Doan
DIFFINT is a reconstruction‑based anomaly detector that uses a differentiable autoencoder with a latent bottleneck composed of soft, axis‑aligned interval memberships. Each latent unit represents a human‑readable hyper‑rectangle in feature space, allowing the model to encode how strongly an instance falls inside each interval and to compute reconstruction error as the anomaly score. The method provides a certified lower bound on reconstruction error for points outside all active intervals, a suppression mechanism for sparse abnormalities, and a closed‑form, label‑free importance ranking for each (unit, feature) pair, achieving top performance on 48 ADBench benchmarks against 22 baselines.
By Lamine Diop, Marc Plantevit
The paper explores how prior-data fitted networks (PFNs) can be adapted for anomaly detection in tabular data. Using frozen TabPFN features, the authors show that scoring samples by nearest-neighbor distance in feature space yields strong results, and they identify optimal layers and extraction procedures. They further improve performance by fine‑tuning the model on a reference set, achieving higher AUROC on the ADBench benchmark with both a fine‑tuning‑free (ZEN) and a fine‑tuned (FOCUS) approach.
DIFFINT is a reconstruction‑based anomaly detector that replaces the opaque latent bottleneck of a standard autoencoder with a set of soft, axis‑aligned interval memberships learned directly from raw numerical data. Each latent unit represents a human‑readable hyper‑rectangle, and an instance’s anomaly score is its reconstruction error weighted by how strongly it falls inside these intervals. The method provides a certified lower bound on reconstruction error for points outside all active intervals, a graded suppression mechanism for sparse anomalies, and a closed‑form, label‑free importance ranking for each (unit, feature) pair, achieving top performance on 48 ADBench benchmarks against 22 baselines.
whyItMatters":"DIFFINT offers the first interpretable anomaly detector that maintains competitive performance while revealing which feature ranges drive each anomaly score, enabling practitioners to audit and understand model decisions without requiring anomaly labels."
arXiv:2611.08136v1 Announce Type: new
Abstract: Anomaly detection (AD) is often framed as a representation-learning problem, where strong performance is expected to require learned features or AD-spe...
By Hyojun Yi, Geunyoung Kim, Hogeon Seo, Seunghyoung Ryu
The paper investigates whether the performance of anomaly detection systems can be predicted without labeled anomalies. For kNN-based detectors, it derives a lower bound on AUC that links detection performance to the separation and variance of inlier and outlier scores, and uses this to analyze how density variation, intrinsic dimensionality, and domain mismatch affect score variability. The authors introduce pseudo‑anomaly probes that provide a reference for estimating relative score separation, and demonstrate through experiments on DCASE benchmarks that these probes enable anomaly‑free model selection to outperform conventional development‑set selection, especially under domain shift.
By Kevin Wilkinghoff, Zheng-Hua Tan
WAND is an unsupervised tabular anomaly detector that scores each point by how far its projection on unit‑sphere directions deviates from a sub‑Gaussian baseline. The directions that flag a point serve as its explanation, providing per‑feature attribution at no extra cost and recoverable via gradients. On 47 ADBench datasets, WAND matches or exceeds 16 baselines in ROC‑AUC while delivering more accurate, faithful explanations than post‑hoc SHAP, LIME, or ECOD, all with linear scoring time and a probe‑efficiency guarantee.
By Lamine Diop
Anomaly detection (AD) is often framed as a representation-learning problem, where strong performance is expected to require learned features or AD-specific optimization. However, this view can overlo...
arXiv:2606. 28970v1 Announce Type: cross Abstract: Unsupervised tabular anomaly detection requires methods that are accurate, robust across heterogeneous datasets, and computationally efficient.
By Quanling Zhao, Jiaying Yang, Ye Tian, Josh Victoria, Zhijun Wang, Pietro Mercati, Onat Gungor, Tajana Rosing
arXiv:2608.29702v1 Announce Type: new
Abstract: A token-embedding table holds a hub of short rows near its origin, and we show that this cluster biases what nearest-neighbor intrinsic-dimension (ID)...
By Alexandre Quemy
Interpretable AI with Local Distillation proposes a method where a black‑box teacher model guides a regularized linear student model at each query point. The teacher defines locality by upweighting training observations with similar predicted outcomes and anchors the fit with its own prediction at the query point, treated as a pseudo‑observation. By adding Gaussian randomization and refitting, the approach identifies reliable features and stable subgroups, achieving near‑teacher accuracy while producing sparse, locally interpretable linear models.
By Erin Craig, Yiling Huang, Snigdha Panigrahi
This paper introduces IDEAL, a framework for few-shot anomaly detection that uses both normal and anomalous reference examples. IDEAL suppresses irrelevant normal variations and encodes intrinsic deviation vectors that capture discriminative anomaly directions. Experiments on eight real-world datasets show that IDEAL generalizes to unseen anomalies and outperforms existing methods.
By Huan Wang, Jun Shen, Jun Yan, Guansong Pang