Recommender systems operate as black boxes, preventing users and regulators from steering their outputs toward specific intentions or auditing their behavior. This lack of controllability, defined as the system's ability to respond to explicit guidance, remains an unaddressed dimension in existing evaluation paradigms. To fill this gap, we propose CtrlBench-Rec, a collaborative multi-agent framework for systematic assessment of controllability. We formalize three fundamental tasks: target content discovery, interest profile shaping, and popularity bias mitigation, which together measure steerability from explicit commands to implicit representation steering and finally to overcoming algorithmic biases.
Extensive experiments on real-world datasets and multiple recommendation models demonstrate that our framework effectively quantifies controllability and exposes critical system bottlenecks, most notably persistent resistance to guiding long tail content. CtrlBench-Rec provides the first standardized toolkit for controllable recommendation research, algorithmic auditing, and user empowerment. Our code is released on GitHub.
Blogger's Review: This paper introduces a crucial framework for assessing controllability in recommender systems, addressing a significant gap in the field. The development of CtrlBench-Rec not only equips researchers with practical tools but also lays the groundwork for enhancing algorithm transparency and user autonomy.