一句话：提出 Bifrost，用自监督对比学习为日志学习多层级故障性表示，提升基于日志的故障诊断。
现有基于日志的故障诊断方法常使用自然语言预训练语言模型（PLM）进行日志表示，但系统故障体现在日志的多层级结构中，PLM 难以全面捕获这些多层级故障信息，即 fallibility representations，因而无法满足故障诊断需求。Bifrost 借鉴站点可靠性工程师的日志分析经验，精心设计基于自监督对比学习的策略来学习日志的故障性表示。在三个公开系统和一个工业 ML-as-a-Service 系统上，Bifrost 产生的日志表示优于现有 PLM：异常检测 F1 平均提升 9.83%，根因定位 HR@k 平均提升 18.28%，故障识别 Macro-F1 平均提升 20.88%。
Log-based fault diagnosis is crucial for runtime debugging and maintenance. Existing fault diagnosis methods use language models pre-trained on natural language (PLMs) for log representation. However, system faults are reflected in the multi-level structure of system logs. PLMs pre-trained on natural language struggle to comprehensively capture multi-level fault information, failing to meet the requirements of fault diagnosis. We refer to this information as fallibility representations. To address this problem, we propose a novel log representation learning method, Bifrost. It draws inspiration from the log analysis experience of Site Reliability Engineers and meticulously designs strategies based on self-supervised contrastive learning to learn the fallibility representations of logs. Across three public systems and one industrial ML-as-a-Service system, the log representations produced by Bifrost outperform existing PLMs by average margins of 9.83% in F1 for anomaly detection, 18.28% in HR@k for root cause localization, and 20.88% in Macro-F1 for fault identification.
@inproceedings{abs-2607-23169,
title = {Bifrost: Empowering Pretrained Language Model with Fallibility Representation for Log-Based Fault Diagnosis},
author = {Minghua He and Tong Jia and Lingzhe Zhang and Chiming Duan and Xinlong Zhao and Leyi Pan and Cheng Wang and Kangjin Wang and Yinghao Yu and Liping Zhang and Yifan Wu and Ying Li},
booktitle = {ASE},
year = {2026},
doi = {10.1145/3832783.3834433}
}