Jan Leike· janleike·· 2026-01-21AI 评分40
2025年主流模型对齐度显著提升
Interesting trend: models have been getting a lot more aligned over the course of 2025. The fraction of misaligned behavior found by auto...
AI 导读
有趣的趋势：在 2025 年期间，模型的对齐程度大幅提升。自动化审计发现的未对齐行为比例不仅在 Anthropic 下降，GDM 和 OpenAI 也是如此。
来源：Jan Leike · x.com