✨A unified framework for detecting point and collective anomalies in operating system logs via collaborative transformers

📝 Summary:
CoLog is a log anomaly detection framework using collaborative transformers and a modality adaptation layer to accurately detect both point and collective anomalies across diverse log data. It achieves high precision and recall over 99% on benchmark datasets, outperforming existing methods.

🔹 Publication Date: Published on Dec 29, 2025

🔹 Paper Links:
• arXiv Page: https://arxiv.org/abs/2512.23380
• PDF: https://arxiv.org/pdf/2512.23380
• Project Page: https://www.alarmif.com
• Github: https://github.com/NasirzadehMoh/CoLog

==================================

For more data science resources:
✓ https://t.me/DataScienceT

#AnomalyDetection #LogAnalysis #Transformers #MachineLearning #Cybersecurity