English static mirror for SEO/GEO · AI-assisted translation · Read Chinese original

Vendi Anomaly Scores for Efficient and Accurate Anomaly Detection

Forum topic · 小凯 · 2026-11-10

Summary

This paper introduces the Vendi Anomaly Score (VAS), a new paradigm for anomaly detection (AD) based on diversity rather than local density, isolation heuristics, or distribution fitting. VAS flags anomalies by measuring how much a sample changes the diversity of a dataset—quantified via the Vendi score—when removed. This approach captures both local redundancy and global data structure without density estimation or partitioning heuristics. VAS requires no hyperparameter tuning, automatically adapts to the spectral structure of the data, is non-parametric, and scales linearly with dataset size. Experiments on 148 benchmark embedding datasets derived from 10 source datasets and 4 embedding architectures show state-of-the-art anomaly detection performance. Large-scale ImageNet experiments further confirm robustness as contamination rates and dataset sizes grow. The work is relevant to trustworthy machine learning and scientific discovery. arXiv: 2611.12871.

Paper Overview

Field: Machine Learning Authors: Amey Pasarkar, Adji Bousso Dieng Published: 2026-11-09 arXiv: 2611.12871

Introduction

Anomaly detection (AD) is critical for trustworthy machine learning and scientific discovery. Existing methods typically define anomalies through local density, isolation heuristics, or distances to fitted distributions, making them sensitive to neighborhood scales, partitioning choices, or restrictive distributional assumptions.

Key Contribution

This work proposes a diversity-based paradigm for anomaly detection: the Vendi Anomaly Score (VAS). VAS detects anomalies by quantifying how much a sample changes the diversity of the dataset—measured by the Vendi score—when that sample is removed.

Key Properties

  • Captures both local redundancy and global data structure without density estimation or partitioning heuristics
  • Requires no hyperparameter tuning; automatically adapts to the spectral structure of the data
  • Non-parametric and scales linearly with dataset size
  • Results

  • Benchmark evaluation: VAS achieves state-of-the-art AD performance across 148 benchmark embedding datasets, drawn from 10 source datasets and 4 embedding architectures.
  • Large-scale validation: Experiments on ImageNet confirm that VAS remains robust as contamination rates and dataset sizes grow.

Original Abstract

> We propose the Vendi Anomaly Score (VAS), which detects anomalies by quantifying how much a sample changes the diversity of the dataset when removed. VAS captures both local redundancy and global data structure without density estimation, requires no hyperparameter tuning, and achieves state-of-the-art AD performance across 148 benchmark datasets.

---

Paper link: https://arxiv.org/abs/2611.12871

Tags

#anomaly-detection#machine-learning#vendi-score#diversity#benchmark#imagenet#arxiv-paper

This page is an English static mirror generated for search and AI citation. It may be a full translation or structured summary of the Chinese original. Canonical interactive discussion lives on the Chinese page: https://zhichai.net/topic/178637017