Shuai Liu

Hello, I'm

Shuai Liu

Founding Team Member @ Synvo AI

Multimodality Egocentric AI Agentic Memory

01About

I am currently a Founding Team Member at Synvo AI. Previously, I was a Core Contributor at LMMs-Lab, where I was on an exciting journey towards LMMs and feeling the AGI.

Prior to this, I obtained my MEng (by Research) degree at MMLab@NTU, where I was fortunate to be advised by Prof. Ziwei Liu and to have received kind and valuable guidance from Jingkang Yang and Bo Li. I obtained my bachelor's degree in Artificial Intelligence from BUPT in 2024.

Additionally, I had a wonderful time as an intern at the Shanghai AI Laboratory under the mentorship of Prof. Ziwei Liu, and I also completed internships at NLPR@CASIA and HAOMO.AI.

My research lies in Multimodality, Egocentric AI, and Agentic Memory.

Feel free to reach out for collaborations, questions, or just to chat!

02News

2026

2025

  • Apr 29 Aero-1-Audio: Our first generation of lightweight audio models, outperforming Whisper and Qwen-2-Audio.
  • Mar 01 EgoLife is accepted by CVPR 2025.
  • Jan 23 LMMs-Eval is accepted by NAACL 2025 Findings.
  • Jan 01 Joined Synvo AI as a Founding Team Member.

2024

  • Aug 13 Joined MMLab@NTU as a master student.
  • Jul 17 Released LMMs-Eval, a comprehensive benchmark for evaluating Large Multimodal Models.
  • Jul 01 Octopus is accepted by ECCV 2024.
  • Jun 12 Released lmms-eval/v0.2.0 to support video evaluations.

2023

  • Oct 12 Introduced Octopus, an embodied vision-language programmer that plays GTA-V.
  • Aug 20 PSG4D is accepted as NeurIPS 2023 Spotlight.

03Selected Publications

FileGram 2026

FileGram: Grounding Agent Personalization in File-System Behavioral Traces

Shuai Liu, Shulin Tian, Kairui Hu, Yuhao Dong, Zhe Yang, Bo Li, Jingkang Yang, Chen Change Loy, Ziwei Liu

Demo-ICL arXiv 2026

Demo-ICL: In-Context Learning for Procedural Video Knowledge Acquisition

Yuhao Dong, Shulin Tian, Shuai Liu, Shuangrui Ding, Yuhang Zang, Xiaoyi Dong, Yuhang Cao, Jiaqi Wang, Ziwei Liu

EgoLife CVPR 2025

EgoLife: Towards Egocentric Life Assistant

Jingkang Yang, Shuai Liu, Hongming Guo, Yuhao Dong, et al.

LMMs-Eval NAACL 2025

LMMs-Eval: Reality Check on the Evaluation of Large Multimodal Models

Kaichen Zhang, Bo Li, Peiyuan Zhang, Fanyi Pu, Joshua Adrian Cahyono, Kairui Hu, Shuai Liu, Yuanhan Zhang, Jingkang Yang, Chunyuan Li, Ziwei Liu

Octopus ECCV 2024

Octopus: Embodied Vision-Language Programmer from Environmental Feedback

Jingkang Yang*, Yuhao Dong*, Shuai Liu*, Bo Li*, Ziyue Wang, Chencheng Jiang, Haoran Tan, Jiamu Kang, Yuanhan Zhang, Kaiyang Zhou, Ziwei Liu

PSG4D NeurIPS 2023

4D Panoptic Scene Graph Generation

Jingkang Yang, Jun Cen, Wenxuan Peng, Shuai Liu, Fangzhou Hong, Xiangtai Li, Kaiyang Zhou, Qifeng Chen, Ziwei Liu

NTU MEng Thesis

Egocentric Perception and Memory: From Embodied Understanding to a Lifelong Assistant

Shuai Liu — Nanyang Technological University, 2025

04Projects

Engram Suite

Engram Suite

Privacy-first AI memory layer — "Signal for AI Memory." End-to-end encrypted memory system for AI assistants with zero-knowledge sync.

E2E Encryption MCP Local-first
Cocoa Suite

Cocoa Suite

A local AI assistant that transforms files into actionable memory through semantic understanding. Fully private, on-device, zero cloud dependency.

On-device Multimodal Vector Search