I’m Ben Chen (陈犇), Now is the group lead of post‑training of multimodal foundation models at Kuaishou Technology. Before 26/08, I were the group leader of AI Search Team. I have developed OneSearch, the first end-to-end generative retrieval framework for search engine, as well as Search-E1, a self-evolving search agent system. Now I am working on Multimodal Reasoning, Agentic RL, Agentic Search (Mutli-Agent), and Recursive Self-Improvement (RSI).

I have published more than 50 papers at the top international AI conferences, including Search-E1, OneSearch, InfoGain-RAG, and BianQue. I have been awarded the best paper award and outstanding paper award as the first author in the last three years. The total google scholar citations are about 1000+.

If you are seeking any form of academic or work collaboration, please feel free to email benchen4395@gmail.com.

📖 Employments

  • 2024.06 - Now, Multimodal Retreival and AI Search, Kuaishou Technology.
  • 2020.07 - 2024.06, International Commerce Business Unit, Alibaba Group.

🔥 News

📝 Publications

NeruIPS 2026
sym

OneSearch-V2: The Latent Reasoning Enhanced Self-distillation Generative Search Framework
Ben Chen, Siyuan Wang, Yufei Ma, Zihan Liang, Xuxin Zhang, Yue Lv, Ying Yang, Huangyu Dai, Lingtao Mao, Tong Zhao, Zhipeng Qian, et. al. (NeurIPS 2026)

  • A latent reasoning enhanced self-distillation generative search framework. Codes and data are public on OneSearch-Family
  • It effectively mitigates common issues such as information bubbles and long-tail sparsity, without incurring additional inference costs or serving latency
ICML 2026
sym

OneSearch: A Preliminary Exploration of the Unified End-to-End Generative Framework for E-commerce Search
Ben Chen, Xian Guo, Siyuan Wang, Zihan Liang, Yue Lv, Yufei Ma, Xinlong Xiao, Bowen Xue, Xuxin Zhang, Ying Yang, Huangyu Dai, et. al. (ICML 2026)

  • The first industrial-deployed end-2-end generative framework for e-commerce search, with 24.06% gains in MFU and 75.40% reduction in OPEX.
  • Various offline and online A/B tests are conducted, verifying its effectiveness and efficiency for the real e-commerce search scenarios.

📖 Educations

  • 2017.09 - 2020.06, Master, Huazhong University of Science and Technology, Wuhan.
  • 2013.09 - 2017.06, Undergraduate, Wuhan University, Wuhan.

🎖 Honors and Awards

  • 2024.05, The Outstanding paper of ECML-PKDD 2024.
  • 2021.02, The Best paper of AAAI2021 Shared Task Track.

💬 Invited Talks

💻 Internships

  • 2019.07 - 2020.06, Alibaba, Hangzhou.