Add design spec for Humans vs LLMs: Humanity's Last Exam
Commit
elizabethsiegle1mo ago
Documents the Cloudflare Workers architecture, HLE question curation approach (gated dataset baked into a static bundle rather than fetched live), the simulated model-comparison stat, and the durable leaderboard design agreed on during brainstorming.
Co-Authored-By: Claude Sonnet 5 noreply@anthropic.com Entire-Checkpoint: 22c154dbc097