{"ID":2890247,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2507.20051","arxiv_id":"2507.20051","title":"$K^4$: Online Log Anomaly Detection Via Unsupervised Typicality Learning","abstract":"Existing Log Anomaly Detection (LogAD) methods are often slow, dependent on error-prone parsing, and use unrealistic evaluation protocols. We introduce $K^4$, an unsupervised and parser-independent framework for high-performance online detection. $K^4$ transforms arbitrary log embeddings into compact four-dimensional descriptors (Precision, Recall, Density, Coverage) using efficient k-nearest neighbor (k-NN) statistics. These descriptors enable lightweight detectors to accurately score anomalies without retraining. Using a more realistic online evaluation protocol, $K^4$ sets a new state-of-the-art (AUROC: 0.995-0.999), outperforming baselines by large margins while being orders of magnitude faster, with training under 4 seconds and inference as low as 4 $μ$s.","short_abstract":"Existing Log Anomaly Detection (LogAD) methods are often slow, dependent on error-prone parsing, and use unrealistic evaluation protocols. We introduce $K^4$, an unsupervised and parser-independent framework for high-performance online detection. $K^4$ transforms arbitrary log embeddings into compact four-dimensional d...","url_abs":"https://arxiv.org/abs/2507.20051","url_pdf":"https://arxiv.org/pdf/2507.20051v1","authors":"[\"Weicong Chen\",\"Vikash Singh\",\"Zahra Rahmani\",\"Debargha Ganguly\",\"Mohsen Hariri\",\"Vipin Chaudhary\"]","published":"2025-07-26T20:24:51Z","proceeding":"cs.LG","tasks":"[\"cs.LG\",\"cs.CL\",\"cs.DC\"]","methods":"[]","has_code":false}