GRAB: An LLM-Inspired Sequence-First Click-Through Rate Prediction Modeling Paradigm
This addresses the problem of improving ad recommendation accuracy and revenue for platforms like Baidu, though it appears incremental as it builds on LLM-inspired scaling for a specific domain.
The paper tackles the performance and efficiency bottlenecks of traditional Deep Learning Recommendation Models (DLRMs) by proposing GRAB, an end-to-end generative framework for Click-Through Rate (CTR) prediction inspired by Large Language Models (LLMs), resulting in a 3.05% increase in revenue and a 3.49% rise in CTR in online deployment.
Traditional Deep Learning Recommendation Models (DLRMs) face increasing bottlenecks in performance and efficiency, often struggling with generalization and long-sequence modeling. Inspired by the scaling success of Large Language Models (LLMs), we propose Generative Ranking for Ads at Baidu (GRAB), an end-to-end generative framework for Click-Through Rate (CTR) prediction. GRAB integrates a novel Causal Action-aware Multi-channel Attention (CamA) mechanism to effectively capture temporal dynamics and specific action signals within user behavior sequences. Full-scale online deployment demonstrates that GRAB significantly outperforms established DLRMs, delivering a 3.05% increase in revenue and a 3.49% rise in CTR. Furthermore, the model demonstrates desirable scaling behavior: its expressive power shows a monotonic and approximately linear improvement as longer interaction sequences are utilized.