Reproduce “HELP: Hierarchical Embeddings-based Log Parsing” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Andy Xu, Arno Gau
HELP parses streaming software logs by clustering them with tuned text embeddings and sending newly formed clusters to an LLM for template extraction. It periodically merges similar clusters to address shifts. Across 14 datasets, it averages FGA of 0.896 versus Drain’s 0.550, but remains behind offline LILAC overall. This supports an online benchmark advantage, not superiority over all parsers.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.