2026/07/26 by Srivaths Ranganathan, Srikanth Ranganathan, Zihuan Diao +9
Computer Science · #Gaussian Processes and Bayesian Inference #Music and Audio Processing #Recommender Systems and Techniques #cs.IR
paper · pdf · doi:10.1145/3773078.3831876
openalex publication_date 2026/07/26 · openalex created_date 2026/07/30 · openalex updated_date 2026/07/31
Continuously trained ranking models in music recommenders fall into feedback loops where previously consumed items dominate recommendations. This suppresses two distinct content classes: new releases (temporal freshness) and unlistened catalog items (novelty). Industry practitioners have a wide menu of interventions available, ranging from serving-time heuristics, training-data reweighting, architectural debiasing, to uncertainty-driven exploration, each of which are well understood in academic settings. But live systems offer challenges with continuously ingested content, interconnected components, and practical limitations that counteract the findings from academic research. We report results from off-policy online A/B tests for six interventions and a combination experiment across four conceptual layers (serving, training, architecture, exploration) on the YouTube Music homepage. All interventions modify the ranking model or the serving layer that consumes its scores; candidate generation and other upstream components are held fixed. We discuss key takeaways from our results: first, serving-time interventions on continuously trained systems are neutralized by the learning loop. Second, architectural debiasing reduces popularity dominance and improves diversity but does not create discovery, while carrying hidden integration costs. Finally, uncertainty-driven exploration interventions with a Spectral-normalized Neural Gaussian Process (SNGP) head produce the largest new-release lift, though they come with a measurable engagement or diversity tradeoff. We close with recommendations on which layer to intervene at, and the hidden costs of each choice.