Author

Lin-Jun Li

1 paper indexed here

We haven’t gathered this author’s papers yet. Follow them and we’ll fetch their work.

Not the right person? Other researchers publish under this name.

#large language models Jul 2026

Same Dangerous Objective, Opposite Advice: Direct Exposure versus Multi-Agent Mediation

A current high-capability model can be used as the user-facing component of an automated, multi-stage workflow serving an explicitly manipulative objective, and this result exposes a compositional safety gap.

Lin-Jun Li · arXiv.org · 0 citations