Current Work: Mentir-AI

Supporting Teachers. Building on Student Thinking. Advancing Mathematical Learning.

MENTIR-AI (Mathematics Education Network for Teaching and Instructional Responses) explores how artificial intelligence can support teachers in creating opportunities for deep and sustained mathematical engagement and learning.

Students' mathematical ideas, strategies, and questions provide valuable starting points for teaching and learning. Teachers bring the mathematical knowledge, pedagogical expertise, and understanding of their students needed to recognize and build on those ideas.

MENTIR-AI seeks to extend, rather than replace, that expertise by developing AI-supported tools that help teachers interpret student thinking, provide meaningful feedback, plan and facilitate instruction, and engage in ongoing professional learning.

Help Shape MENTIR-AI

We're inviting mathematics teachers to become partners in developing and improving MENTIR-AI.

Teachers' experience and professional judgment are essential to creating tools that are genuinely useful for teaching and learning. We are working with teachers to examine student mathematical work, evaluate AI-generated feedback, and explore how AI might better support classroom instruction and professional growth.

Participating teachers may have opportunities to review and improve AI-generated responses, share their perspectives on instructional challenges, and help test developing tools. We are particularly interested in teachers' different interpretations and approaches, which can reveal important possibilities that AI systems might otherwise overlook.

[Learn About Participation Opportunities]

What Are We Developing?

Our initial focus is on supporting teachers in providing meaningful, individualized feedback on student mathematical work. Rather than simply identifying correct and incorrect answers, we want AI-supported feedback to recognize students' reasoning, build on emerging mathematical ideas, and encourage continued exploration.

Our broader vision includes support for:

• Feedback on student work: Helping teachers recognize and respond to students' mathematical ideas.

• Classroom interactions: Supporting teachers in facilitating discussions and responding to student thinking as it emerges.

• Instructional preparation and reflection: Helping teachers anticipate student strategies, explore mathematical tasks, and reflect on instructional decisions.

• Professional learning: Creating opportunities for teachers to deepen their mathematical and pedagogical knowledge through inquiry and collaboration.

Across these areas, our goal is to develop AI tools that support teachers' professional judgment and expand opportunities for meaningful mathematical engagement.

Building on Decades of Research and Practice

MENTIR-AI builds on more than three decades of work beginning with the **Math Forum**, whose Problems of the Week program engaged students and mentors in conversations about mathematical reasoning. Its extensive archive of student solutions and mentoring interactions provides a foundation for our current work.

Subsequent projects, including EnCoMPASS and ORCIDS, extended this tradition through research and professional learning environments in which teachers collaboratively examine student thinking, engage in mathematical inquiry, and reflect on their instructional practice.

MENTIR-AI brings these traditions together with emerging AI technologies and the expertise of practicing teachers.

The project is led by Drexel University, in collaboration with Ursinus College and the 21st Century Partnership for STEM Education (21PSTEM), with support from the Gates Foundation.