AI RESEARCH

Compositional Steering of Large Language Models with Steering Tokens

arXiv CS.LG

ArXi:2601.05062v2 Announce Type: replace-cross Deploying LLMs in real-world applications requires controllable output that satisfies multiple desiderata at the same time. While existing work extensively addresses LLM steering for a single behavior, \textit{compositional steering} -- i.e., steering LLMs simultaneously towards multiple behaviors -- remains an underexplored problem. In this work, we propose \emph{compositional steering tokens} for multi-behavior steering.