Keiji AI LogoKeiji AI

Research · Patterns 2022

HINT

Research on clinical trial outcome prediction. HINT predicts whether a clinical trial will succeed. It encodes the drug molecule, target disease and eligibility criteria, then connects them in a hierarchical interaction graph.

Who created HINT?

Tianfan Fu is the first author of HINT. Its last author is Jimeng Sun of the University of Illinois Urbana-Champaign, who went on to co-found Keiji AI in 2023 and is its CEO. The work predates the company.

What the paper found

  • Trained and validated on 1,160 phase I, 4,449 phase II and 3,436 phase III trials.
  • F1 scores of 0.665, 0.620 and 0.847 on separate test sets of 627 phase I, 1,653 phase II and 1,140 phase III trials.

How HINT relates to TrialMind

It is research, not a TrialMind module or a separately sold product. Published methods like this one inform how TrialMind works; TrialMind is the product Keiji AI sells.

See how TrialMind handles this work: Competitive and portfolio intelligence.

The paper

Tianfan Fu, Kexin Huang, Cao Xiao, Lucas M. Glass, Jimeng Sun. HINT: Hierarchical interaction network for clinical-trial-outcome predictions. Patterns, 2022.

Details checked against the published record on 2026-11-01.

Frequently asked questions

Who created HINT?
Tianfan Fu is the first author of HINT. Its last author is Jimeng Sun of the University of Illinois Urbana-Champaign, who went on to co-found Keiji AI in 2023 and is its CEO. The work predates the company.
Is HINT part of TrialMind?
No. It is research, not a TrialMind module or a separately sold product. Published methods like this one inform how TrialMind works; TrialMind is the product Keiji AI sells.
What did the HINT paper find?
Trained and validated on 1,160 phase I, 4,449 phase II and 3,436 phase III trials. F1 scores of 0.665, 0.620 and 0.847 on separate test sets of 627 phase I, 1,653 phase II and 1,140 phase III trials.

All publications

Every paper sits under its primary research area.