Action is All You Need: Dual-Flow Generative Ranking Network for Recommendation

TL;DR

This paper introduces the Dual-Flow Generative Ranking Network (DFGR), a novel model that enhances recommendation system efficiency by reducing computational complexity while outperforming existing models on multiple datasets.

Contribution

DFGR employs a dual-flow mechanism to optimize interaction modeling, significantly improving training and inference efficiency over prior generative ranking models.

Findings

DFGR outperforms DLRM, HSTU, and other models on open-source and industrial datasets.

The dual-flow mechanism reduces computational overhead and enhances efficiency.

Optimal parameter strategies are identified for computationally constrained environments.

Abstract

Deep Learning Recommendation Models (DLRMs) often rely on extensive manual feature engineering to improve accuracy and user experience, which increases system complexity and limits scalability of model performance with respect to computational resources. Recently, Meta introduced a generative ranking paradigm based on HSTU block that enables end-to-end learning from raw user behavior sequences and demonstrates scaling law on large datasets that can be regarded as the state-of-the-art (SOTA). However, splitting user behaviors into interleaved item and action information significantly increases the input sequence length, which adversely affects both training and inference efficiency. To address this issue, we propose the Dual-Flow Generative Ranking Network (DFGR), that employs a dual-flow mechanism to optimize interaction modeling, ensuring efficient training and inference through end-to-end token processing. DFGR duplicates the original user behavior sequence into a real flow and a fake flow based on the authenticity of the action information, and then defines a novel interaction method between the real flow and the fake flow within the QKV module of the self-attention mechanism. This design reduces computational overhead and improves both training efficiency and inference performance compared to Meta's HSTU-based model. Experiments on both open-source and real industrial datasets show that DFGR outperforms DLRM, which serves as the industrial online baseline with extensive feature engineering, as well as Meta's HSTU and other common recommendation models such as DIN, DCN, DIEN, and DeepFM. Furthermore, we investigate optimal parameter allocation strategies under computational constraints, establishing DFGR as an efficient and effective next-generation generative ranking paradigm.