LinkedIn Feed enables professionals worldwide to discover relevant content, build connections, and share knowledge at scale. We present Feed Sequential Recommender (Feed-SR), a transformer-based sequential ranking model for LinkedIn Feed that replaces a DCNv2-based ranker and meets strict production constraints. We detail the modeling choices, training techniques, and serving optimizations that enable deployment at LinkedIn scale. Feed-SR is currently the primary member experience on LinkedIn's Feed and shows significant improvements in member engagement (+2.10% time spent) in online A/B tests compared to the existing production model. We also describe our deployment experience with alternative sequential and LLM-based ranking architectures and why Feed-SR provided the best combination of online metrics and production efficiency.
@article{arxiv.2602.12354,
title = {An Industrial-Scale Sequential Recommender for LinkedIn Feed Ranking},
author = {Lars Hertel and Gaurav Srivastava and Syed Ali Naqvi and Satyam Kumar and Yue Zhang and Borja Ocejo and Benjamin Zelditch and Adrian Englhardt and Hailing Cheng and Andy Hu and Antonio Alonso and Daming Li and Siddharth Dangi and Chen Zhu and Mingzhou Zhou and Wanning Li and Tao Huang and Fedor Borisyuk and Ganesh Parameswaran and Birjodh Singh Tiwana and Sriram Sankar and Qing Lan and Julie Choi and Souvik Ghosh},
journal= {arXiv preprint arXiv:2602.12354},
year = {2026}
}