Lab record
Research publications, AI infrastructure milestones, and updates from the lab.
Our lab’s internal model API and token platform has crossed 300 billion tokens and is now named OPL Gateway. As of July 1, 2026, OPL Gateway has served 2,346,924 requests, distributed 304.96 billion tokens, and …
Since our lab’s in-house large language model platform went live on March 12, 2026, it has now delivered more than 100 billion tokens for internal research and daily work. As of April 29, 2026, the platform has …
One Person Lab App is now the published desktop product for One Person Lab. The App repository publishes the macOS package, release assets, updater metadata, first-run guidance, and the one-shot installer.
Since our lab’s in-house large language model platform went live on March 12, 2026, it has now delivered more than 60 billion tokens for internal research and daily work. As of April 18, 2026, the platform has …
Since our lab’s in-house large language model platform went live on March 12, 2026, it has now delivered more than 30 billion tokens for internal research and daily work. As of April 6, 2026, the platform has …
Our lab publicly released MedAutoScience on GitHub on March 29, 2026. It is an agent-first, human-auditable platform for progressing disease-focused medical studies from versioned data assets to …
Our team published INTELCAPE in Clinical Gastroenterology and Hepatology, an AI system for automated capsule endoscopy analysis that supports faster and more accurate Crohn’s disease diagnosis. INTELCAPE …
Our team published MotifCC in Cancer Letters, a multi-omics immune classification framework for colorectal cancer that links tumor heterogeneity to prognosis and immunotherapy opportunities. By integrating …
Our lab’s in-house large language model platform officially went live on March 12, 2026, and is now available to all lab members. The platform provides unlimited access to GPT 5.4 for internal research and …
On March 6, 2026, our lab completed the relocation to our new site at 4th Floor, Building M3, Phase IV, Guangzhou International Bio Island. The move gives the lab a new base for research, engineering, and …
Our lab’s new parallel storage system officially went online on January 21, 2026, adding 3PB of new capacity for data-intensive research workflows. With this upgrade, the lab now operates 5PB of parallel …
Our team published PCsRNAdb in Nucleic Acids Research, a comprehensive pan-cancer resource for studying small noncoding RNAs across tumor types. The database was built to support systematic exploration of …
Our team introduced CRCFound in Advanced Science, a self-supervised colorectal cancer CT foundation model pretrained on large-scale unlabeled 3D CT data. CRCFound was systematically evaluated on multiple …
Our team published SegMamba-V2 in IEEE Transactions on Medical Imaging, a general framework for long-range modeling in 3D medical image segmentation. The model extends Mamba-based sequential modeling into …
Our team published PIANOS in Nature Communications to address a long-standing challenge in colorectal cancer prognosis: model instability across platforms and cohorts. PIANOS is a platform-independent, …
Our team reported Brim in Advanced Science, an interpretable multimodal fusion model for bridged histology-genomics survival prediction in pan-cancer. Brim is trained with both whole-slide pathology images and …
Our team published a review in Genomics Proteomics Bioinformatics on the opportunities and challenges of AI-driven biomedical multimodal data fusion. The review summarizes how deep learning and foundation-model …
Our team reported in International Journal of Surgery that tertiary lymphoid structure heterogeneity can define clinically meaningful colorectal cancer subgroups. The study integrated multi-omics data, digital …
Our team published TMO-Net in Genome Biology, an explainable pretrained multi-omics model designed for multi-task learning in oncology. The work focuses on learning transferable biological representations …
Our team reported a MedComm study showing that senescence-related molecular programs can stratify colorectal cancer into clinically meaningful subtypes. By integrating transcriptomic, genomic, digital …
Our team introduced SOUSA in Medical Image Analysis, a unified weakly and semi-supervised framework for medical image segmentation using sparse annotations. SOUSA is designed for realistic clinical annotation …
Feng, happy Teacher’s Day.
Congratulating chuling for his graduation. And he is continuing studying with us during his PhD. Welcome to Dongwei, Shiqi and Jiaxin joining us as postgraduates. Welcome to Baowen studying with us during his …
Congratulating Qiqi for her PhD graduation.
Our team published a study in Nature Communications showing that longitudinal imaging can be combined with multi-task deep learning to improve treatment response prediction in rectal cancer. The model jointly …
As the executive chairman, Feng successfully presided over this conference with Prof. Wu and Prof. Wang. By online remote video reportings and on-site presentations of professional researchers from various …
After communication and joint learning with Wei’s team, Feng took us to visit Moganshan scenic area and Nanxun Ancient Town.
Min-Er joined our group as a postdoc with PhD training at Peking Union Medical College Hospital in May. Wei-Bin joined our group as a research assistant with extensive programming experience in June. De-Jun …
As the executive chairman, Feng successfully presided over this meeting. With quite a lot professional researchers attending, we have a huge exchange and collision of ideas in this intellectual feast.
Try another keyword or reset the filters.