Cross-domain Semantic IDs from organic feed activity, quantized via RQ-FSQ, improve industrial ads CTR prediction with gains up to +0.351% AUC at 30x smaller storage.
arXiv:2506.16698 [cs.LG] https://arxiv.org/abs/2506.16698 Sohini Roychowdhury, Doris Wang, Qian Ge, Joy Mu, and Srihari Reddy
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
years
2026 2verdicts
UNVERDICTED 2representative citing papers
DUET pre-trains dedicated transformers for click and conversion streams, yielding up to 0.38% NE reduction over baselines in OCVR prediction.
citing papers explorer
-
Quantizing Intent: Cross-Domain Semantic IDs from Organic Activity for Industrial Ranking
Cross-domain Semantic IDs from organic feed activity, quantized via RQ-FSQ, improve industrial ads CTR prediction with gains up to +0.351% AUC at 30x smaller storage.
-
DUET -- Dual User Embedding Transformers for Offsite Conversion Prediction
DUET pre-trains dedicated transformers for click and conversion streams, yielding up to 0.38% NE reduction over baselines in OCVR prediction.