吴光亚
Employers on the program
- 2026得物1 talk
Events1
Conferences1
Talks1
Years active2026
The year around now
Talks in the past 12 months and announced talks ahead, one row per conference, January to December. Hover a mark for the talk.
Appearances by year
- AICon 全球人工智能开发与应用大会 (AICon Global AI Development and Application Conference)
Talks
| Date | Event | Conference | Role | Title | Links |
|---|---|---|---|---|---|
| Jun 26, 2026 | AICon 全球人工智能开发与应用大会 | 上海 | AICon 全球人工智能开发与应用大会 (AICon Global AI Development and Application Conference) | speaker | 得物电商多场景排序大模型：从千万到十亿参数的Scaling Up实践 as 得物 针对得物交易推荐“4大核心场景 + 50余个细分场景”的多层级独特生态，本演讲聚焦大模型时代排序模型的Scaling Up挑战。在剖析业界探索的基础上，重点阐述得物如何构建统一的排序大模型，以应对极致的场景异构性。方案通过跨域联合建模与MoE-Transformer融合架构，系统性解决数据、特征及训练推理难题，实现模型…Show less针对得物交易推荐“4大核心场景 + 50余个细分场景”的多层级独特生态，本演讲聚焦大模型时代排序模型的Scaling Up挑战。在剖析业界探索的基础上，重点阐述得物如何构建统一的排序大模型，以应对极致的场景异构性。方案通过跨域联合建模与MoE-Transformer融合架构，系统性解决数据、特征及训练推理难题，实现模型从千万到十亿级的平滑扩展。最终在显著提升推荐精准度的同时，为得物数十个差异化业务场景带来了整体的成本控制与业务增长。 演讲提纲： 背景与挑战：推荐排序模型的Scaling Up困境 业界突破Scaling Law的探索与启示 得物电商多场景… |
Frequent co-speakers
No one else spoke at two or more of the same events.
Speakers with the same name are merged into one entry, so this page may combine more than one person. Affiliations are shown as each conference listed them.