You need to agree to share your Hugging Face information to access this dataset

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

Log in or Sign Up to review the conditions and access this dataset content.

Company Lookalikes: Validation Set

75 seed companies for developing and tuning against the openbenchmarks.com/lookalikes benchmark. Every provider receives the same seed and returns the companies most similar to it.

Dataset Description

75 seed companies across 12 market categories.

Each row is one seed company:

  • case_id: stable identifier, prefixed val-.
  • seed_slug, seed_name: the seed company's slug and name.
  • seed_domain: the domain sent to providers that take a company as input.
  • query: the natural-language request sent to providers that take text as input.
  • description, short_description: what the seed company sells.
  • anchor: what the seed fundamentally is and who buys it, for example restaurant point-of-sale platform. A lookalike must match it.
  • capabilities: what the seed's product does. A lookalike needs to share at least one.
  • category: one of 12 market categories.

Scoring

Every provider receives the same seed and returns up to 100 companies. A panel of LLM judges labels each returned company relevant or not by majority vote. A company is relevant when it sells the same kind of product as the seed and its buyers overlap with the seed's buyers. Company size, geography, funding, and maturity are not relevance requirements. Customers of the seed's category, parent or holding companies, agencies or integrators that serve the category, companies in an unrelated industry, and the seed itself are not relevant.

The headline metrics are precision at 10, 25, and 100: the share of the top K returned companies that are relevant, averaged over seeds. A provider that errors or returns nothing for a seed scores N/A on that seed, not zero.

Downloads last month
14