01
asddsad
adadda
People. Narratives. Technology.
Advancing Explainable AI for a Transparent Future.
We develop AI systems that people can understand, question, and trust—bridging foundational research with meaningful human impact.
SPEMJAMI
4 researchers
working across 3 institutions
Transparent
intelligence
ACEInterpretabilityHuman trustintelligence
Scroll to discover
05+Peer-reviewed publications
03+Active research projects
04+Researchers & collaborators
03+Partner institutions
What we explore
Research with purpose.
Four interconnected areas. One shared goal: making intelligent systems more legible, responsible, and useful.
02
Explainable AI
Building models that communicate their reasoning clearly, faithfully, and responsibly.
03
Natural Language
Advancing language understanding that is robust, inclusive, and grounded in human values.
04
Computer Vision
Helping machines interpret visual worlds with transparency and contextual awareness.
05
Advance AI
Explain How AI works
06
Human–AI Interaction
Designing systems that augment human capability while preserving agency and trust.
Inside the lab
Ideas in motion.
From new evaluation frameworks to human-centered tools, our projects turn rigorous inquiry into tangible progress.
Ongoing01Advance AI
Research Lab
Ongoing02Explainable AI
Prism: Interpretable Multimodal Models
Ongoing03Human–AI
ClarityBench: Evaluating AI Explanations
Selected work
Latest publications.
Explainable AI
MI, TA, Nath
Faithful by Design: Rethinking Explanations in Foundation Models
M. Chen, A. Okafor, S. Patel
When Humans Disagree with AI: A Framework for Calibrated Reliance
E. Morgan, J. Alvarez, M. Chen
VisGround: Language-Guided Concept Discovery in Vision Models
S. Patel, T. Wu, A. Okafor
Beyond the journal
More ways to explore PNT.
Discover long-form scholarship, protected innovations, independent coverage, and conversations with our researchers.
The people behind the workDifferent perspectives.
Different perspectives.
Shared curiosity.
Our interdisciplinary team brings together computer scientists, cognitive researchers, designers, and domain experts.
SP
Samir Patel
PhD Researcher
EM
Elena Morgan
Research Scientist
JA
Javier Alvarez
PhD Researcher
MI
M I AYUB
Lead
From the lab
Notes, news & ideas.
Paper Accepted
May 28, 2025
What makes an explanation useful? It depends who is asking.
New findings challenge one-size-fits-all approaches to model explanations.
PNT researchers at CHI 2025 in Yokohama
Welcoming four new researchers to PNT
Work with us
Interested in collaborating or joining us?
Whether you’re looking to collaborate, pursue research, or support our mission—we’d like to hear from you.