Who Holds #1 on Every Hugging Face Official Benchmark? A Live Index

说明与用途

创建了一个实时索引，聚合所有 Hugging Face 官方基准的排行榜，显示哪个组织在各领域占据最多的第一名。

问题

Hugging Face official benchmark leaderboards are spread across dozens of dataset pages, making it hard to identify which organization holds the most first-place spots across all benchmarks.

结果

The index tracks ~49 official benchmarks with ~1,000 leaderboard entries, showing that FINAL‑Bench holds the most first‑place spots (eleven) across various fields.

工作流

• 1. Query Hugging Face API for list of datasets tagged as official benchmarks.
• 2. For each dataset, pull the ranked entries from the public API.
• 3. Aggregate results: count #1 positions per organization, per country, per field, compute gap to #2, and list top two models per benchmark.
• 4. Render a live view with charts, heatmaps, and tables.
• 5. Update automatically on each page visit.
• 6. Handle ties by counting each tied organization.
• 7. Derive country from publishing organization, not model name.
• 8. Measure #1‑to‑#2 gap relative to #2 score and direction‑aware for lower‑is‑better metrics.

证据

Hugging Face marks certain datasets as official benchmarks. Each one carries a leaderboard on its dataset page, built automatically from the .eval_results files that model repositories publish.

来源

The problem: these leaderboards live on dozens of separate dataset pages. There is no single place that answers a simple question.

来源

So we built one.

来源

It is a live, read-only view that recomputes from the Hugging Face API on every visit.

来源

Nothing is hand-entered. The page reads the list of datasets that Hugging Face tags as official benchmarks, pulls each board's ranked entries from the public API, and aggregates.

来源

As of early October 2026 the index tracked around 49 official benchmarks with roughly a thousand leaderboard entries in total, and the single organization holding the most first-place spots was FINAL-Bench with eleven...

来源