newsOfficialPublished: 1d ago

Improving instruction hierarchy in frontier LLMs

IH-Challenge trains models to prioritize trusted instructions, improving instruction hierarchy, safety steerability, and resistance to prompt injection attacks.

OpenAI Source: OpenAI

Why it matters

Company Updates is moving the AI conversation right now, and this update helps explain what changed.

Social and community

Post to X

Permalink

Improving instruction hierarchy in frontier LLMs

Why it matters

Social and community

Discussion