/new Binxin Gao | Researcher & Builder
Binxin Gao - Stylized Portrait

Binxin Gao 高彬馨

I study the economics of AI and innovation. Part of my work is funded by OpenAI Researcher Access Program.

PhD Student at UMD

Former Investor / Game Designer / Social Influencer

View Research Get in Touch
Scroll
01

News

Mar 2026

One paper on LLM math reasoning has been accepted to ICLR'26!

Mar 2026

Excited to present my ongoing work "Buy to build, buy to bluff?" at WINDS (Wharton Innovation Doctoral Symposium) 2026!

Jan 2026

Honored to receive Google Cloud Academic Research Grant!

Dec 2025

Honored to receive Google TPU Research Cloud Program grant!

May 2025

Honored to receive OpenAI Researcher Access Program grant!

02

A Curious Path

Pursuing net positive impact, driven by curiosity.

I study how innovation ecosystems evolve, bridging quantitative finance, technology product development, and venture capital experience.

Before academia, I spent years in industry—designing games at NetEase, building AI products at Tencent, and investing in technology ventures at CCB International. This journey was never linear, but always driven by one question: How do we create products/systems that generate net positive impact?

Now, as a PhD student at the University of Maryland's Robert H. Smith School of Business, I study the economics of AI and innovation.

Beyond research, I'm a Hanfu influencer with 60k+ followers, an award-winning photographer (1st prize among 1.1M+ entries), and an occasional calligrapher.

03

Ongoing Projects

🏆 OpenAI Researcher Access Program 2025

AI-Augmented Innovation in Game Development

Awarded exclusive API access by OpenAI to explore how humans collaborate with AI to innovate. Games serve as the ideal testing ground for frontier technology.

  • Investigating human-AI co-creation workflows
  • Games as laboratories for AI capability exploration
  • Funded research on innovation dynamics
Working Paper

OpenAI or CloseAI? That is A Question

  • Modeling the dynamics of openness in AI development
  • Empirical evidence on open vs. proprietary strategies
  • How firms navigate innovation tensions
Cultural Venture

Hanfu Brand & Dunhuang Collaboration

  • Founded a Hanfu brand with 60k+ followers
  • Collaborated with Dunhuang Research Institute
  • 16 original heritage-inspired designs
04

Publications

2026

Max It or Miss It: Benchmarking LLM On Solving Extremal Problems

Gao, B., & Han, J.

International Conference on Learning Representations (ICLR) 2026

arXiv
2025

Buy to Build or Buy to Bluff? Patent Acquisitions and Startup IPO Performance

Gao, B.

Working Paper

Using 1,007 VC-backed IPOs merged with USPTO patent assignments, I find that startups purchasing patents in the 12 months before IPO experience 10 percentage points lower first-day returns than non-buyers. Exploiting within-firm timing variation, I show markets penalize only last-minute acquisitions while rewarding gradual accumulation, with a 10% increase in pre-IPO purchases reducing closing prices by 0.8%. These results suggest sophisticated investors distinguish genuine capability building from window dressing based on acquisition timing.

2015

Non-linear Momentum Strategies: Theories and Applications in the Chinese Future Market

Gao, B., Li, C., Qiao, K.

Quarterly Journal of Finance, 9(1), pp. 23–58

1 citation Recognized for tenure requirements at top Chinese universities
PDF DOI
05

Community Contributions

Reviewer

  • NeurIPS (Conference on Neural Information Processing Systems) 2025

Talks

  • Wharton Innovation Doctoral Symposium (WINDS), 2026
06

Reflections

Notes on research, technology, and the creative process.

Coming soon...

07

Let's Connect