FlowTime introduces continuous generative regression using a one-step VAE and normalizing flows for personalized priors to predict watch time while addressing mean-collapse, quantization, and latency issues in prior paradigms.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
FlowTime: Towards Continuous Generative Watch Time Prediction via Flow-based Personalized Priors
FlowTime introduces continuous generative regression using a one-step VAE and normalizing flows for personalized priors to predict watch time while addressing mean-collapse, quantization, and latency issues in prior paradigms.