LogCopilot is an LLM framework that builds a hierarchical knowledge base from logs and generates/executes LogQL queries from natural language instructions, reporting 76.8% average accuracy across four datasets.
Dhole, Ishan Sharma, Avani Wildani, José Cambronero, and Andreas Züfle
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.SE 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
LogCopilot: Automating Log Aggregation Analysis through Large Language Models
LogCopilot is an LLM framework that builds a hierarchical knowledge base from logs and generates/executes LogQL queries from natural language instructions, reporting 76.8% average accuracy across four datasets.