用循环结构提升点击率预测模型性能，训练时多循环、推理时零循环。
LoopCTR: Unlocking the Loop Scaling Power for Click-Through Rate Prediction
- 通过循环复用共享层实现计算增长，参数不随规模线性增加。
- 单次前向传播即超越所有基线，工业部署更高效。
- 适合追求高精度且受限于推理资源的推荐系统场景。
通过堆叠更多参数来扩展基于Transformer的点击率（CTR）模型会带来日益增长的计算与存储开销，导致扩展理想与工业部署限制之间的差距不断扩大。我们提出LoopCTR，引入一种循环扩展范式，通过递归复用共享模型层增加训练时计算量，将计算与参数增长解耦。LoopCTR采用增强型三明治架构，结合超连接残差与专家混合机制，并在每个循环深度施加过程监督，将多循环优势编码至共享参数中。该方法支持‘训练多循环、推理零循环’策略，单次前向传播即超越所有基线。在三个公开基准和一个工业数据集上的实验表明其达到顶尖性能。奥数分析进一步揭示0.02–0.04 AUC的未开发潜力，训练时循环更少的模型具有更高的奥数上限，指明了自适应推理的广阔前景。
原文摘要 · Abstract (English)
Scaling Transformer-based click-through rate (CTR) models by stacking more parameters brings growing computational and storage overhead, creating a widening gap between scaling ambitions and the stringent industrial deployment constraints. We propose LoopCTR, which introduces a loop scaling paradigm that increases training-time computation through recursive reuse of shared model layers, decoupling computation from parameter growth. LoopCTR adopts a sandwich architecture enhanced with Hyper-Connected Residuals and Mixture-of-Experts, and employs process supervision at every loop depth to encode multi-loop benefits into the shared parameters. This enables a train-multi-loop, infer-zero-loop strategy where a single forward pass without any loop already outperforms all baselines. Experiments on three public benchmarks and one industrial dataset demonstrate state-of-the-art performance. Oracle analysis further reveals 0.02--0.04 AUC of untapped headroom, with models trained with fewer loops exhibiting higher oracle ceilings, pointing to a promising frontier for adaptive inference.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。