Safe Context Switching for Agents in the Wild: Mitigating Subspace Interference via Orthogonal Adaptation

Excerpt

arXiv:2610.05219v1 Announce Type: new Abstract: Most Large Language Models exhibit a fundamental tension between two sequential tasks, such as logical reasoning and safety alignment. The high-variance internal states required for sophisticated Chain-of-Thought (CoT) deduction can geometrically interfere with latent representations encoding safety constraints. We identify this phenomenon as Sequential Subspace Interference, showing that standard fine-tuning on logical tasks such as multi-step mat