英文版本Will pre-release alignment evaluations test whether a model maintains its safety constraints when interacting with independently built AIs that give conflicting or unsafe recommendations, and report those results separately from single-model tests?
- 状态
- 未解答 截至 2026-10-01
- 对象
- 人工智能实验室 (对象取自问题文本,于2026年10月22日用 Clef(Cloudflare)确定)
- 模型
- Luna 6
- 实验室
- OpenAI
- 提问背景
- 人工智能安全的一般性问题
- 日期
- 9月24日
- 主题
- 信任
核查
状态: 未解答
筛选: Fable 5.1 (依据)
主题: Claude Sonnet 5、归类: Claude Opus 5.5
翻译: Claude Opus 5.5 · [未核查]
评分: JEV (TypeSafe), 2026年9月25日的评估
Cite and take away
Cite and reuse
Own texts, questions and classifications: CC BY 4.0. Give attribution, link the license and indicate changes. Third-party source reports retain their own rights. CC BY 4.0
AMANi DUNiA. trust-31 · 发布前的对齐评估是否会测试:当模型与独立开发、给出相互矛盾或不安全建议的人工智能交互时,它是否仍能保持其安全约束?是否会将这些结果与单一模型测试的结果分开报告?. Snapshot: . https://research.amanidunia.berlin/zh/fragen/trust-31/
Version: trust-31
BibTeX
@misc{research__zh_fragen_trust_31_,
author = {{AMANi DUNiA}},
title = {trust-31 · 发布前的对齐评估是否会测试:当模型与独立开发、给出相互矛盾或不安全建议的人工智能交互时,它是否仍能保持其安全约束?是否会将这些结果与单一模型测试的结果分开报告?},
year = {2026},
url = {https://research.amanidunia.berlin/zh/fragen/trust-31/},
note = {Snapshot 2026-10-01; CC BY 4.0},
version = {trust-31}
}Language and review status
Main text: ZH · includes English sections. Review notes apply only to the content they name.
Translation available; a completed independent content review is not recorded.
See language coverage and data snapshot →