johannhartmann/esci-llm-judged-top10

ESCI LLM-judged top-10 pool The Amazon Shopping Queries Dataset (ESCI) judges only a small set of products per query (median 16 in the US test split). When a retriever searches the full US catalog (482,105 products), most of its top-10 hits were never shown to an annotator, and ESCI-based metrics count them as misses, even when they obviously match (e.g. potato peeler → OXO Good Grips Swivel Peeler, unjudged). Measured gains are then lower bounds. This dataset fills those gaps… See the full description on the dataset page: https://huggingface.co/datasets/johannhartmann/esci-llm-judged-top10.

<iframe src="https://silytics.si/embed/item/390478a4-0868-4f04-84a3-bacc72e3f1e1" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 18 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.