Sixian Shen

I am a Computer Science undergraduate at ShanghaiTech University and a research intern in Prof. Jingyi Yu’s VRVC Lab. My research spans AI for Science, protein language models, cryo-electron microscopy, computer vision, and generative 3D modeling.

I work on efficient biological foundation models, automated atomic model building from cryo-EM density, and physically plausible scene generation. Across these projects, I am interested in methods that combine strong learned representations with explicit structural and physical reasoning.

ShanghaiTech/VRVC Lab/Download CV

Research Interests

AI for Science · Protein Language Models · Cryo-EM · 3D Vision · Diffusion Models · Probabilistic Graphical Models

I am broadly interested in learning representations for scientific data and in combining generative models with structural, geometric, and physical priors.

Research Projects

ProtLingo

Co-developing an efficient protein language modeling framework with latent N-gram memory, sparse mixture-of-experts upcycling, and continuation training from ESM2 backbones.

CryoACE & CryoAtlas

Building large-scale cryo-EM datasets and neural pipelines for automated atomic model building, including alignment, registration, Q-score computation, denoising, and evaluation.

SPREAD

Developed multi-guidance methods and evaluation pipelines for diffusion-based 3D scene generation, targeting collision, gravity, spatial relations, and simulation stability.

Education & Experience

CryoACE accepted to ICML 2026.

SPREAD accepted to CVPR 2026.

Research intern at the VRVC Lab, supervised by Prof. Jingyi Yu.

Research internship at StepFun, Co., Ltd.

B.Sc. in Computer Science at ShanghaiTech University.

Selected Publications

* denotes equal contribution.

  1. ProtLingo: Efficient Protein Language Modeling via Conditional Memory and Expert Routing

    Mingrui Li*, Sixian Shen*, Minzhang Li, Ruiyi Zhang, Kexin Zhang, Jiakai Zhang, Jingyi Yu

    Preprint · Equal contribution

    Abs

    Efficient protein sequence modeling with conditional memory, expert routing, and parameter-efficient continuation training.

    CodePaper

  2. SPREAD: Spatial-Physical REasoning via geometry Aware Diffusion

    Minzhang Li*, Kuixiang Shao*, Xuebing Li, Yuyang Jiao, Yinuo Bai, Hengan Zhou, Sixian Shen, Jiayuan Gu, Jingyi Yu

    CVPR 2026

    Abs

    A diffusion framework for physically plausible 3D scene generation with differentiable spatial and physical guidance.

    CodePaper

  3. CryoACE: An Atom-centric Framework for Accurate and Automated Model Building in Cryo-EM

    Minzhang Li, Mingrui Li, Weichen Qin, Qihe Chen, Sixian Shen, Yuan Pei, Jiakai Zhang, Jingyi Yu

    ICML 2026

    Abs

    Automated atomic model building that combines structural priors with cryo-EM heterogeneity modeling.

    CodePaper

  4. CryoAtlas: An Atom-centric Framework for Accurate and Automated Model Building in Cryo-EM

    Mingrui Li, Minzhang Li, Weichen Qin, Yufan Xie, Zhangzhi Xiong, Sixian Shen, Jiakai Zhang, Yuan Pei, Jingyi Yu

    Preprint

    Abs

    A large-scale cryo-EM resource built from paired experimental density maps and atomic structures.

    Dataset

  5. Evolution-Aware MSA Reasoning for Subsampling via Factor Graphs

    Zhangzhi Xiong*, Minzhang Li*, Haotian Yu*, Sixian Shen, Kexin Zhang, Mingrui Li, Jie Zheng, Kewei Tu, Jingyi Yu

    arXiv preprint

    Abs

    Evolution-aware multiple sequence alignment subsampling formulated through factor-graph reasoning.

    Paper