PhD Student · University of Toronto
Sheza Munir
Conflict resolution, collective intelligence, and consensus in AI
I am a PhD student in Computer Science at the University of Toronto, advised by Dr. Syed Ishtiaque Ahmed in the Third Space Lab. I study how conflicts get resolved between AI models, between humans, and between people and machines, and how collective intelligence and consensus can be built without silencing minority views. I develop and evaluate reasoning-based methods for reaching shared outcomes, and ask when disagreement should be resolved and when it should be preserved.
I did my undergraduate degree in Computer Science at LUMS, advised by Dr. Agha Ali Raza, and my Master's at the University of Michigan, where I worked with Dr. Lu Wang on LLM factuality evaluation. I am a Massey College Junior Fellow, a Research Fellow at the Collective Intelligence Project, and co-lead of WisdomBench with the MIT Global Humanities Initiative.
News
- Oct 8, 2026Posted the preprint Activating the Annotator on arXiv.
- Oct 5, 2026Passed my first PhD milestone!
- Sep 2026Co-leading WisdomBench with the MIT Global Humanities Initiative, benchmarking AI across wisdom cultures.
- Sep 2026Became a Massey College Junior Fellow.
- Aug 2026Became a Research Fellow at the Collective Intelligence Project.
- Jun 25–28, 2026The Consensus Trap presented at FAccT 2026, Montréal. Paper ↗
- Jun 13–17, 2026Attended HCIC 2026.
- May 25–29, 2026Presented A Fair, Multi-Perspective Ethical Reasoning Framework at Canadian AI 2026, Burnaby.
- Apr 23–27, 2026ExpertLongBench at ICLR 2026, Rio de Janeiro. Paper ↗
- Nov 5–9, 2025Verifact at EMNLP 2025, Suzhou. Paper ↗
- Aug 2025Started my PhD at the University of Toronto.
- Jul 27–Aug 1, 2025FactBench at ACL 2025, Vienna. Paper ↗
- Aug 11–16, 2024Presented Deepfake Defense, an Urdu deepfake audio dataset, at ACL Findings 2024, Bangkok. Paper ↗
Latest newsScroll for older news ↓
Research Interests
Conflict Resolution
How disagreements between AI models, between people, and between the two get resolved, and when they should be preserved instead.
Collective Intelligence
Mechanisms for building consensus from many voices: reasoning-based aggregation, social choice, and what each method keeps or erases.
Pluralistic Evaluation
Benchmarks such as WisdomBench that evaluate AI across wisdom cultures and perspectives instead of a single right answer.
Evaluation & Safety
Measuring the reliability of language models, from long-form factuality to deepfake detection in low-resource languages.
Publications
2026
Preprint · Under review
Activating the Annotator: Operationalising Lived Experience in Data Annotation
2026
FAccT 2026
The Consensus Trap: Dissecting Subjectivity and the "Ground Truth" Illusion in Data Annotation
2026
CAI 2026
Are You the A-hole? A Fair, Multi-Perspective Ethical Reasoning Framework
2026
ICLR 2026
ExpertLongBench: Benchmarking Language Models on Expert-Level Long-Form Generation Tasks with Structured Checklists
2025
ACL 2025
FactBench: A Dynamic Benchmark for In-the-Wild Language Model Factuality Evaluation
2025
EMNLP 2025
Verifact: Enhancing Long-Form Factuality Evaluation with Refined Fact Extraction and Reference Facts
2024
ACL Findings 2024
Deepfake Defense: Constructing and Evaluating a Specialized Urdu Deepfake Audio Dataset
All publicationsScroll for more ↓
Experience & Education
Sep 2026 – Present
Junior Fellow
Massey College, University of Toronto
Sep 2026 – Present
Co-lead, WisdomBench
MIT Global Humanities Initiative
Evaluating and benchmarking AI on different wisdom cultures.
Aug 2026
Research Fellow
Collective Intelligence Project
Aug 2025 – Present
PhD in Computer Science
University of Toronto · Third Space Lab · Advisor: Dr. Syed Ishtiaque Ahmed
Studying how conflicts between people, between AI models, and between the two can be resolved, and how consensus can be reached without erasing minority views. Developing and evaluating explanation-based and formal reasoning approaches. Passed first PhD milestone (Oct 2026).
Jun 2024 – May 2025
Research Assistant, LAUNCH Lab
University of Michigan, Ann Arbor · Advisor: Dr. Lu Wang
Developed VERIFY, a factuality evaluation pipeline for LLMs, and FACTBENCH, a 1,000-prompt benchmark for identifying hallucination triggers. Paper accepted at ACL 2025.
2023 – May 2025
Master of Health Information (Data Science and Machine Learning)
University of Michigan, Ann Arbor
100% merit scholarship (full tuition).
Jan 2022 – Jul 2023
Machine Learning Intern, AI in Healthcare Group
Lahore University of Management Sciences (LUMS)
Built LSTM models on wearable time-series data to predict migraines 24 hours in advance with 88% accuracy. Student as Co-Researcher Grant, 2022.
Aug 2021 – May 2023
Research Assistant & Lab Head, CSALT
Lahore University of Management Sciences (LUMS)
Pioneered a 35,000-audio Urdu deepfake detection dataset. Managed a team of 33 students across data acquisition, annotation, and model training. First paper from Pakistan accepted at ACL (Findings 2024).
2019 – May 2023
BSc in Computer Science
Lahore University of Management Sciences (LUMS) · Advisor: Dr. Agha Ali Raza
Silver Medal, GPA 3.97/4.0. Dean's Honor List, rank 2 of 242.
Teaching
Teaching Assistant, CSC 300: Computers and Society
University of Toronto · Aug 2025 – present · Facilitating discussion on the ethical, legal, and sociotechnical implications of pervasive AI
Instructor, SIADS 642: Deep Learning
University of Michigan · Winter & Summer 2024 · Mentored 100+ graduate students in deep neural architectures
Graduate Student Instructor, SI 504: Git, Shell and Servers
University of Michigan · Aug 2023 – May 2025 · Mentored 300+ students in version control and shell scripting
Teaching Assistant, CS 433 & CS 466
LUMS · Fall 2021, Winter 2022 · ML for Speech Processing; Human Computer Interaction
Awards & Honours
Fellowships & Funding
2026
Massey College Junior Fellowship
University of Toronto
2026
Research Fellowship
Collective Intelligence Project
2023–25
100% Merit Scholarship
University of Michigan · Full tuition for the Master's program
2022
Student as Co-Researcher (ScR) Grant
LUMS · Awarded for the Migraines Prediction Project
Awards & Competitions
2023 & 2024
Outstanding Graduate Student Instructor Award, Nominee
University of Michigan · SI 504: Git, Shell and Servers
2023
Winner, Health Tech Pitch Competition
Aga Khan University · Health Data for Pakistan conference · Migraines Prediction Project
2022
Best Project, Aarzu Physical Therapy App
LUMS Collaborative Project Showcase
2019
International Chemistry Olympiad, National Winner
51st IChO · Represented Pakistan · Global Honourable Mention, top finish among 80 countries
2016
Outstanding Volunteer Service Award
U.S. Department of State
Beyond Research
Leadership & Service
Social Coordinator, CS Graduate Society
University of Toronto · 2026
Co-Chair, CS Academy for High School Students
University of Toronto · 2026
GAAP Mentor, Graduate Application Assistance Program
University of Toronto · 2025
Chapter Lead, IEEE LUMS Women in Computing
LUMS · 2021–2022
VP, Innovation & Design Society
LUMS · 2021–2022
Google Developer Student Club Lead
LUMS · 2021–2022
Selected Projects
Aarzu: Physical Therapy App
mHealth app for low-income, low-literate users at Pakistan Society for Rehabilitation of the Disabled. 100% task-completion with doctors, 84.4% with patients.
Migraines Prediction Engine
LSTM models on wearable time-series data (heart rate, EDA, stress) for 5 patients. Predicted migraines one day in advance with 88% accuracy. ScR Grant.
MLH Open Source Fellowship
Major League Hacking · Summer 2022 · Built unit testing and search-engine failure backup systems in Python
Contact
I'm always open to new collaborations and conversations about conflict
resolution between AI and humans, collective intelligence, and consensus
building, especially with researchers working in social choice, negotiation,
fairness, and AI safety.