An LLM-BiLSTM hybrid model detects anomalous HDFS log blocks with 94.8% F1 on the Loghub HDFS dataset, outperforming DeepLog.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Exploring Block Anomaly Detection In HDFS Log Data Analysis
An LLM-BiLSTM hybrid model detects anomalous HDFS log blocks with 94.8% F1 on the Loghub HDFS dataset, outperforming DeepLog.