You're viewing this list as a guest. Sign in to comment or star items.Sign in
A competition to get small / old models to perform at frontier levels through better prompting. Score = (benchmark perf) / (model size or compute cost).tmad4000.github.io
Thoughts on cross-domain identity systems. Browser extension vs private GitHub repo + PAT vs encrypted-content-in-public-repo etc.tmad4000.github.io
See-also reference from the brainstorms index. AI-discovered research opportunities.tmad4000.github.io