METR, an AI safety research organization, detailed its 2023 accomplishments, including developing methodologies for evaluating AI agents on autonomous tasks and contributing to OpenAI's GPT-4 system card. The organization also proposed "Responsible Scaling Policies" (RSPs), a framework for AI safety that gained traction among researchers and companies like Anthropic and OpenAI. Additionally, METR partnered with the UK AI Safety Institute and evaluated GPT-5.1 for catastrophic risks. AI
排序理由 METR's year-in-review details research and evaluation methodologies, including contributions to a system card and a proposed safety framework that saw industry adoption.
在 METR (Model Evaluation & Threat Research) 阅读 →
- Anthropic
- Eric Schmidt
- Geoffrey Hinton
- GPT-4
- GPT-5.1
- OpenAI
- Responsible Scaling Policies
- UK AI Safety Institute
- White House Executive Order on AI
- Yoshua Bengio
AI 生成摘要 · Google Gemini · 来自 5 个来源。 我们如何撰写摘要 →
