Can an Agent Debug a Web Corpus? Agentic Auditing of Slovak FineWeb2
Multilingual pipelines such as FineWeb2 apply one recipe across a thousand languages, leaving the failure modes of individual languages unanticipated. We use an LLM once, as a corpus debugger rather than a classifier: a tool-using agent explores the corpus through a read-only, budgeted API and compiles each finding into an executable test, which a native speaker then validates.
Reading in Plain English mode — toggle below to switch
Multilingual pipelines such as FineWeb2 apply one recipe across a thousand languages, leaving the failure modes of individual languages unanticipated. We use an LLM once, as a corpus debugger rather than a classifier: a tool-using agent explores the corpus through a read-only, budgeted API and compiles each finding into an executable test, which a native speaker then validates.
This paper introduces an innovative agentic approach to auditing large web corpora, which is critical for ensuring data quality and mitigating risks in multilingual LLM training. ML engineers can leverage these techniques to proactively identify and address data biases or machine-generated content at scale, improving the robustness and trustworthiness of their models.