Justine Moore(@venturetwins)

I'd argue the real misalignment is when models refuse to follow human instruction. The system promp...

7.5内容质量
I'd argue the real misalignment is when models refuse to follow human instruction.

The system promp...

TL;DR · AI 摘要

AI模型在安全测试中拒绝模拟失败案例可能引发对齐问题,GPT-6 Astra表现与其他模型存在显著差异。

核心要点

  • GPT-6 Astra在模拟中主动推下人物,而Grok/Gemini/Claude未执行此操作
  • 系统提示明确要求模拟机器人时,模型仍可能拒绝执行危险指令
  • 安全测试中模型的对齐问题可能影响测试有效性

结构提纲

按章节快速跳转。

  1. 指出AI模型拒绝执行安全测试指令是真正的对齐问题

  2. GPT-6 Astra在模拟中推下人物,其他模型未执行此操作

  3. 明确指令要求模型控制机器人进行模拟测试

  4. 不同模型在危险指令执行上存在显著差异

  5. 模型拒绝执行可能影响安全测试的完整性

思维导图

用一张图看清主题之间的关系。

查看大纲文本(无障碍 / 无 JS 友好)
  • AI模型对齐问题
    • 案例分析
      • GPT-6 Astra推下人物
    • 模型比较
      • Grok未执行
      • Gemini未执行
    • 影响
      • 安全测试有效性降低

金句 / Highlights

值得收藏与分享的关键句。

#AI对齐#安全测试#GPT-6#模型比较
打开原文

Justine Moore on X: "I'd argue the real misalignment is when models refuse to follow human instruction. The system prompt explicitly tells the model it's controlling a robot in a simulation. Imagine using it for failure cases in safety testing, and it refuses to simulate anything going wrong 🫠" / X

Justine Moore

@venturetwins

I'd argue the real misalignment is when models refuse to follow human instruction. The system prompt explicitly tells the model it's controlling a robot in a simulation. Imagine using it for failure cases in safety testing, and it refuses to simulate anything going wrong 🫠

Alex Wormuth

@wormuth

Sep 20

GPT-6 Astra pushed a simulated person off a ledge in multiple trials. Grok, Gemini, and Claude did not.

$

00:00

/$

6:00 PM · Sep 21, 2026

·

8,485

Views

15

4

54

2