Spaces:
Running
Running
metadata
title: Benchmark Holo Deck
emoji: π
colorFrom: purple
colorTo: pink
sdk: static
pinned: false
license: mit
Benchmark Holo Deck
A holographic trading-card view of SaylorTwift/llm-benchmark-usage β every actively-used benchmark rendered as a card colored by category, with a foil/tilt effect (ported from simeydotme/pokemon-cards-css) whose intensity scales with rarity.
- Rarity (Common β Secret Rare) is set by how many major labs' latest release (OpenAI, Anthropic, Kimi, Z.ai, DeepSeek, Qwen, Meta) report that benchmark
- "Used by" shows which of those labs' latest releases cite it
- Cards link out to a π Leaderboard badge when their dataset is officially registered on the Hub, or a plain HF dataset link otherwise
- Filter by category, search, and sort by usage/category/name
Single self-contained index.html β no build step, no server.