GADE Union
a Global AI Data Experts Union

GADE Research

GADE Research is the current strategic priority of GADE Union, focusing on frontier academic research.

We specialize in Multimodal and Agent-based evaluation benchmarks. Unlike conventional datasets, we aim to build high-quality benchmarks that precisely capture model boundaries and critical flaws. We believe that identifying where models "fail" is essential for providing clear direction for the research and iteration of frontier models.

Selected Work

Pix2Fact (2026): An expert-level visual reasoning benchmark project designed to identify factual reasoning flaws in complex Visual Question Answering (VQA) models. Paper Link

Selected Publications

Research

Pix2Fact (2026)

An expert-level visual reasoning benchmark project designed to identify factual reasoning flaws in complex Visual Question Answering (VQA) models. The core findings are currently being submitted to ICML 2026 with GADE members as lead authors

Contact

Join the Network

If you are a Ph.D. expert or a student with exceptional research potential, we welcome you to join us:

Deep Involvement: Join GADE Research to participate in building frontier datasets and high-end expert annotation projects.
Paid Compensation: We provide paid compensation to members for their professional contributions to our projects.
Academic Achievement: High-potential student collaborators have the opportunity to serve as core members in community-led submissions to top-tier academic conferences.
How to Apply: Fill the form on the right side and send your CV/Portfolio to: research@gadeunion.com

Partnerships

We appreciate the interest from various sectors. To ensure research depth and neutrality, our current operational model is as follows:

Focused Priority: Due to current resource allocation, GADE Union is concentrating all core efforts on frontier academic research and open-source benchmark development within GADE Research.
Commercial Status: We are currently not accepting any paid B2B service requests. Our priority is to solidify technical foundations in multimodal and agent-based domains.
Academic Synergy: While we do not undertake commercial projects, we welcome non-profit collaborations with academic foresight (e.g., joint development of open-source benchmarks, academic resource sharing).
Inquiry Email: For academic collaboration interests, please contact: partnerships@gadeunion.com

GADE Uniona Global AI Data Experts Union