mlsanityData to 7 Oct 2026

APG: Adaptive Parameter Generation Network for Click-Through Rate Prediction

arXiv 2203.16218 ↗First posted 30 Mar 2022cs.IR cs.AI cs.LGcited by 12

Not on the map. Why?

Abstract (arXiv)

In many web applications, deep learning-based CTR prediction models (deep CTR models for short) are widely adopted. Traditional deep CTR models learn patterns in a static manner, i.e., the network parameters are the same across all the instances. However, such a manner can hardly characterize each of the instances which may have different underlying distributions. It actually limits the representation power of deep CTR models, leading to sub-optimal results. In this paper, we propose an efficient, effective, and universal module, named as Adaptive Parameter Generation network (APG), which can dynamically generate parameters for deep CTR models on-the-fly based on different instances. Extensive experimental evaluation results show that APG can be applied to a variety of deep CTR models and significantly improve their performance. Meanwhile, APG can reduce the time cost by 38.7% and memory usage by 96.6% compared to a regular deep CTR model. We have deployed APG in the industrial sponsored search system and achieved 3% CTR gain and 1% RPM gain respectively.

Key facts. Cited by 12 papers in the map (data through 7 Oct 2026). Builds on 32 of its 45 references that are in the map. First posted on arXiv 30 Mar 2022.

Builds on

32 of 45 references.

References · 32

  1. 2016HyperNetworkscited by 1,003
  2. 2016Wide & Deep Learning for Recommender Systemscited by 745
  3. 2017DeepFM: A Factorization-Machine based Neural Network for CTR Predictioncited by 673
  4. 2017Learning Efficient Convolutional Networks through Network Slimmingcited by 673
  5. 2020Intrinsic Dimensionality Explains the Effectiveness of Language Model Fine-Tuningcited by 658
  6. 2017Deep Interest Network for Click-Through Rate Predictioncited by 541
  7. 2016Cross-stitch Networks for Multi-task Learningcited by 466
  8. 2017Deep & Cross Network for Ad Click Predictionscited by 392
  9. 2018Measuring the Intrinsic Dimension of Objective Landscapescited by 364
  10. 2018AutoInt: Automatic Feature Interaction Learning via Self-Attentive Neural Networkscited by 319
  11. 2016Dynamic Filter Networkscited by 318
  12. 2021Parameter-efficient Multi-task Fine-tuning for Transformers via Shared Hypernetworkscited by 278
20 more
  1. 2019ACNet: Strengthening the Kernel Skeletons for Powerful CNN via Asymmetric Convolution Blockscited by 178
  2. 2020Search-based User Interest Modeling with Lifelong Sequential Behavior Data for Click-Through Rate Predictioncited by 176
  3. 2015Conditional Computation in Neural Networks for faster modelscited by 166
  4. 2016Product-based Neural Networks for User Response Predictioncited by 145
  5. 2019Knowledge Graph Convolutional Networks for Recommender Systemscited by 121
  6. 2019MeLU: Meta-Learned User Preference Estimator for Cold-Start Recommendationcited by 119
  7. 2021Diverse Branch Block: Building a Convolution as an Inception-like Unitcited by 87
  8. 2019FiBiNET: Combining Feature Importance and Bilinear feature Interaction for Click-Through Rate Predictioncited by 83
  9. 2021One Model to Serve All: Star Topology Adaptive Recommender for Multi-Domain CTR Predictioncited by 78
  10. 2018Contextual Parameter Generation for Universal Neural Machine Translationcited by 40
  11. 2019Neural Input Search for Large Scale Recommendation Modelscited by 39
  12. 2020Towards Automated Neural Interaction Discovery for Click-Through Rate Predictioncited by 37
  13. 2021Deep Learning for Click-Through Rate Estimationcited by 33
  14. 2014Exponentially Increasing the Capacity-to-Computation Ratio for Conditional Computation in Deep Learningcited by 24
  15. 2019IntentGC: a Scalable Graph Convolution Framework Fusing Heterogeneous Information for Recommendationcited by 19
  16. 2019S2DNAS:Transforming Static CNN Model for Dynamic Inference via Neural Architecture Searchcited by 14
  17. 2021Click-Through Rate Prediction with Multi-Modal Hypergraphscited by 12
  18. 2018Decouple Learning for Parameterized Image Operatorscited by 11
  19. 2021Adaptive Convolutions with Per-pixel Dynamic Filter Atomcited by 11
  20. 2021Binary Code based Hash Embedding for Web-scale Applicationscited by 11

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 4; 2023: 4; 2024: 2; 2025: 0; 2026: 2.

12 papers.

  1. 2022AdaSparse: Learning Adaptively Sparse Structures for Multi-Domain Click-Through Rate Predictioncited by 7
  2. 2022Higher-order Neural Additive Models: An Interpretable Machine Learning Model with Feature Interactionscited by 7
  3. 2024Hate Speech Detection with Generalizable Target-aware Fairnesscited by 7
  4. 2022BASM: A Bottom-up Adaptive Spatiotemporal Model for Online Food Ordering Servicecited by 5
  5. 2023Hybrid Contrastive Constraints for Multi-Scenario Ad Rankingcited by 4
  6. 2024GPRec: Bi-level User Modeling for Deep Recommenderscited by 3
  7. 2022DIGMN: Dynamic Intent Guided Meta Network for Differentiated User Engagement Forecasting in Online Professional Social Platformscited by 1
  8. 2023FAN: Fatigue-Aware Network for Click-Through Rate Prediction in E-commerce Recommendationcited by 1
  9. 2023DynInt: Dynamic Interaction Modeling for Large-scale Click-Through Rate Predictioncited by 0
  10. 2023DPAN: Dynamic Preference-based and Attribute-aware Network for Relevant Recommendationscited by 0
  11. 2026UME: A Unified Meta-Generalization Framework for Cross-Domain ETAcited by 0
  12. 2026PRIME: Mitigating Subgroup Optimization Competition in Shared CTR Top Networks with Plug-in Residual Input-Conditioned Mixture of Expertcited by 0