Fu Li
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Fu Li has a total of 29 co-authors. Here are the top co-authors by number of shared publications:
- Jiaqi Zhai
Meta Platforms, Inc., Menlo Park, California, USA
2 shared publications - Xing Liu
Meta Recommendation Systems, Menlo Park, CA, USA
2 shared publications - Yueming Wang
Meta Platforms, Inc., Menlo Park, California, USA
2 shared publications - Zhaojie Gong
Meta Platforms, Inc., Menlo Park, California, USA
2 shared publications - Bin Wen
OpenAI, San Francisco, CA, USA
1 shared publication - Guo Liang
Meta Platforms, Inc., Menlo Park, California, USA
1 shared publication
Latest Publications
- Open Access
- 19 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 4616-4621
Request-Only Optimization for Recommendation Systems
- 19 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 4616-4621
Open AccessRequest-Only Optimization for Recommendation Systems
- Open Access
- 4 August 2023
- Published by Association for Computing Machinery (ACM)
- p. 5520-5531
Revisiting Neural Retrieval on Accelerators
Retrieval finds a small number of relevant candidates from a large corpus for information retrieval and recommendation applications. A key component of retrieval is to model (user, item) similarity, which is commonly represented as the dot product of two learned embeddings. This formulation permits efficient inference, commonly known as Maximum Inner Product Search (MIPS). Despite its popularity, dot products cannot capture complex user-item interactions, which are multifaceted and likely high rank. We hence examine non-dot-product retrieval settings on accelerators, and propose mixture of logits (MoL), which models (user, item) similarity as an adaptive composition of elementary similarity functions. This new formulation is expressive, capable of modeling high rank (user, item) interactions, and further generalizes to the long tail. When combined with a hierarchical retrieval strategy, h-indexer, we are able to scale up MoL to 100M corpus on a single GPU with latency comparable to MIPS baselines. On public datasets, our approach leads to uplifts of up to 77.3% in hit rate (HR). Experiments on a large recommendation surface at Meta showed strong metric gains and reduced popularity bias, validating the proposed approach's performance and improved generalization.
- 4 August 2023
- Published by Association for Computing Machinery (ACM)
- p. 5520-5531
Open AccessRevisiting Neural Retrieval on Accelerators
Retrieval finds a small number of relevant candidates from a large corpus for information retrieval and recommendation applications. A key component of retrieval is to model (user, item) similarity, which is commonly represented as the dot product of two learned embeddings. This formulation permits efficient inference, commonly known as Maximum Inner Product Search (MIPS). Despite its popularity, dot products cannot capture complex user-item interactions, which are multifaceted and likely high rank. We hence examine non-dot-product retrieval settings on accelerators, and propose mixture of logits (MoL), which models (user, item) similarity as an adaptive composition of elementary similarity functions. This new formulation is expressive, capable of modeling high rank (user, item) interactions, and further generalizes to the long tail. When combined with a hierarchical retrieval strategy, h-indexer, we are able to scale up MoL to 100M corpus on a single GPU with latency comparable to MIPS baselines. On public datasets, our approach leads to uplifts of up to 77.3% in hit rate (HR). Experiments on a large recommendation surface at Meta showed strong metric gains and reduced popularity bias, validating the proposed approach's performance and improved generalization.