MATH · IN · MODELS

Orthogonalized personality directions still bleed behaviorally

measured in 1 paper

Bhandari, Naseem & Nasim extract five Big-Five personality steering directions in LLaMA-3-8B-Instruct and Ministral-8B-Instruct and find substantial pairwise cosine overlap [bhandari-etal-2026-personality-traits-interfere-steering] A battery of exact-orthogonalization schemes drives pairwise cosine similarity to near machine precision (max|cos|<1e-8 under hard orthonormalization) [bhandari-etal-2026-personality-traits-interfere-steering] Even then, steering one trait at a time measurably shifts other geometrically orthogonal traits' scores [bhandari-etal-2026-personality-traits-interfere-steering] Geometric orthogonality is therefore not sufficient for behavioral independence, a competing claim against papers finding natural or restorable independence [bhandari-etal-2026-personality-traits-interfere-steering]

Context

personality steering, cosine similarity, orthogonalization, behavioral cross-talk, competing claim

Papers

Do Personality Traits Interfere? Geometric Limitations of Steering in Large Language Models — Bhandari, Pranav, Naseem, Usman, Nasim, Mehwish2026 · arXiv:2602.15847