热点
"可纠正性" 相关文章
Edge Cases in AI Alignment
少点错误 2025-03-24T09:32:10.000000Z
Request for Comments on AI-related Prediction Market Ideas
少点错误 2025-03-02T21:06:56.000000Z
Instrumental Goals Are A Different And Friendlier Kind Of Thing Than Terminal Goals
少点错误 2025-01-24T20:20:36.000000Z
Why Modelling Multi-Objective Homeostasis Is Essential for AI Alignment (And How It Helps With AI Safety as Well)
少点错误 2025-01-12T04:45:14.000000Z
Corrigibility should be an AI's Only Goal
少点错误 2024-12-29T20:35:50.000000Z
Corrigibility's Desirability is Timing-Sensitive
少点错误 2024-12-26T22:34:10.000000Z
Why Worry About Incorrigible Claude?
Astral Codex Ten 2024-12-24T09:09:48.000000Z