Expand More, Shrink Less: Shaping Effective-Rank Dynamics for Dense Scaling in Recommendation

ArXi:2605.23191v1 Announce Type: new Scaling recommendation models is a central challenge in recommender systems. Recently, RankMixer has emerged as an effective solution, operating on a unified token representation and alternating between token mixing and per-token feedforward networks (P-FFNs) to achieve scalable performance. However, RankMixer suffers from \textit{embedding collapse}, where learned representations have low effective rank, limiting expressivity and underutilizing the expanded representation space.