Research

I started in low-resource NLP, then moved into machine learning for cancer care and finance, where a wrong prediction can cause real harm. Working there showed me how often models fail without warning, which led to my current focus: confidence estimation, building methods that let language, reasoning, and vision-language models tell when their own answers are likely wrong.

All24 Confidence & ReliabilityConfidenceEstimators that tell when language, reasoning, and vision-language models are likely wrong, and where they break.8 LLMs for OncologyOncologyPrivacy-preserving LLMs that extract and grade treatment toxicities from clinical notes, with Henry Ford Health.8 Earlier WorkEarlier WorkLow-resource NLP and applied machine learning in finance, sports, and health signals.8

Confidence & Reliability · 8x

New estimators across model families

Grounded or Guessing? LVLM Confidence Estimation via Blind-Image Contrastive Ranking
Reza Khanmohammadi, Erfan Miahi, Simerjot Kaur, Charese H. Smiley, Ivan Brugere, Kundan Thind, Mohammad M. Ghassemi
AACL-IJCNLP 2026Main conferenceCore CS
How Reliable are Confidence Estimators for Large Reasoning Models? A Systematic Benchmark on High-Stakes Domains
Reza Khanmohammadi, Erfan Miahi, Simerjot Kaur, Charese H. Smiley, Ivan Brugere, Kundan Thind, Mohammad M. Ghassemi
EACL 2026Main conferenceCore CS
Calibrating LLM Confidence by Probing Perturbed Representation Stability
Reza Khanmohammadi, Erfan Miahi, Mehrsa Mardikoraem, Simerjot Kaur, Ivan Brugere, Charese H. Smiley, Kundan Thind, Mohammad M. Ghassemi
EMNLP 2025Main conferenceCore CSOutstanding Paper nominee, top 0.4%
In the media:

Stress-testing confidence

Vision-Language Model Confidence Is Not a Property of the Answer
Reza Khanmohammadi, Erfan Miahi, Ivan Brugere, Simerjot Kaur, Charese H. Smiley, Kundan Thind, Mohammad M. Ghassemi
2026PreprintCore CSUnder review
- arXiv
- Code

Confidence where errors are costly

Calibrated Triage, Not Autonomy: Confidence Estimation for Medical Vision-Language Models
Reza Khanmohammadi, Kundan Thind, Mohammad M. Ghassemi
2026PreprintMedicalUnder review
- arXiv
Confidence Estimation for Financial Vision-Language Models in Chart and Document Understanding
Reza Khanmohammadi, Simerjot Kaur, Charese H. Smiley, Ivan Brugere, Mohammad M. Ghassemi
2026PreprintFinanceUnder review
- arXiv
Risk-Controlled Selective Automation of Oncology Chart Review via Representational Stability Calibration
Reza Khanmohammadi, Ahmed I. Ghanem, S. Siddiqui, Hassan Bagher-Ebadian, Benjamin Movsas, Mohammad M. Ghassemi, Kundan Thind
ASTRO 2026AbstractOncologySpotlight
- Paper
Teaching Language Models Safe Clinical Decision Making: Risk-Informed Training to Reduce Catastrophic Errors in Oncology
Reza Khanmohammadi, Ahmed I. Ghanem, Y. Jee, A.R. Bhatnagar, S. Siddiqui, Mohamed A. Elshaikh, Hassan Bagher-Ebadian, Benjamin Movsas, Mohammad M. Ghassemi, Kundan Thind
ASTRO 2026AbstractOncologyOral, top 11%
- Paper

LLMs for Oncology · 8x

Toxicity extraction and grading

Clinical Validation of a Privacy-Preserving LLM for End-to-End CTCAE Toxicity Assessment in Prostate Cancer Radiotherapy
Reza Khanmohammadi, Ahmed I. Ghanem, A.R. Bhatnagar, S. Siddiqui, Mohamed A. Elshaikh, Hassan Bagher-Ebadian, Benjamin Movsas, Mohammad M. Ghassemi, Kundan Thind
ASTRO 2026AbstractOncology
- Paper
Efficient CTCAE Grading for Post-Radiotherapy Toxicities Using Large Language Models: A Privacy-Preserving Approach Using Instruction Fine-Tuning
Reza Khanmohammadi, Ahmed I. Ghanem, A.R. Bhatnagar, Jawad Turfa, S. Siddiqui, Mohamed A. Elshaikh, Hassan Bagher-Ebadian, Benjamin Movsas, Indrin J. Chetty, Mohammad M. Ghassemi, Kundan Thind
ASTRO 2025AbstractOncology
- Paper
Hybrid Student-Teacher Large Language Model Refinement for Cancer Toxicity Symptom Extraction
Reza Khanmohammadi, Ahmed I. Ghanem, Kyle Verdecchia, Ryan D. Hall, Mohamed A. Elshaikh, Bing Luo, Benjamin Movsas, Hassan Bagher-Ebadian, Indrin J. Chetty, Tuka Alhanai, Kundan Thind, Mohammad M. Ghassemi
IEEE BHI 2025ConferenceOncology
- Paper
Advancing Post-Radiotherapy Toxicity Extraction: A Novel Privacy-Preserving, Parameter-Efficient Language Model Fine-Tuning
Reza Khanmohammadi, Ahmed I. Ghanem, Jawad Turfa, M. Salim Siddiqui, Mohamed A. Elshaikh, Hassan Bagher-Ebadian, Benjamin Movsas, Indrin J. Chetty, Mohammad M. Ghassemi, Kundan Thind
AAPM 2025AbstractOncology
- Paper
A Novel Localized Student-Teacher LLM for Enhanced Toxicity Extraction in Radiation Oncology
Reza Khanmohammadi, Ahmed I. Ghanem, Kyle Verdecchia, Ryan D. Hall, Mohamed A. Elshaikh, Benjamin Movsas, Hassan Bagher-Ebadian, Indrin J. Chetty, Mohammad M. Ghassemi, Kundan Thind
ASTRO 2024AbstractOncology
- Paper
Iterative Prompt Refinement for Radiation Oncology Symptom Extraction Using Teacher-Student Large Language Models
Reza Khanmohammadi, Ahmed I. Ghanem, Kyle Verdecchia, Ryan D. Hall, Mohamed A. Elshaikh, Benjamin Movsas, Hassan Bagher-Ebadian, Indrin J. Chetty, Mohammad M. Ghassemi, Kundan Thind
ICCR 2024ConferenceOncologyOral
- Paper
- arXiv
A Novel Learning Approach for Training Transformer-Based Natural Language Processing Pipeline to Extract Toxicity Symptoms in Prostate Cancer Patients after Definitive Radiotherapy
Reza Khanmohammadi et al.
AAPM 2023AbstractOncology

Guides

Integrating natural language processing into radiation oncology: a practical guide to transformer architecture and large language models
Reza Khanmohammadi, Mohammad M. Ghassemi, Kyle Verdecchia, Ahmed I. Ghanem, Bing Luo, Indrin J. Chetty, Hassan Bagher-Ebadian, Farzan Siddiqui, Mohamed A. Elshaikh, Benjamin Movsas, Kundan Thind
BJR|Artificial Intelligence 2025JournalOncology
- Paper
- arXiv

Earlier Work · 8x

Low-resource NLP

Prose2Poem: The Blessing of Transformers in Translating Prose to Persian Poetry
Reza Khanmohammadi, Mitra Sadat Mirshafiee, Yazdan Rezaee Jouryabi, Seyed Abolghasem Mirroshandel
ACM TALLIP 2023JournalCore CS
- Paper
PGST: A Persian gender style transfer method
Reza Khanmohammadi, Seyed Abolghasem Mirroshandel
Natural Language Engineering 2023JournalCore CS
- Paper
- arXiv
COPER: a Query-Adaptable Semantics-based Search Engine for Persian COVID-19 Articles
Reza Khanmohammadi, Mitra Sadat Mirshafiee, Mehdi Allahyari
ICWR 2021ConferenceCore CS
- Paper

Applied machine learning

Bridging Scientific Research, Innovation, and Finance: A Temporal Heterogeneous Graph Dataset for Financial Investment Prediction
Reza Khanmohammadi, K. Singh, P. Maheshwari, V. Panda, Simerjot Kaur, Ivan Brugere, Charese H. Smiley, Armineh Nourbakhsh, Tuka Alhanai, Mohammad M. Ghassemi
PreprintFinanceUnder review
Investigating the Temporal Association of Biomedical Research on Small Business Funding: A Bibliometric and Data Analytic Approach
Reza Khanmohammadi, Simerjot Kaur, Charese H. Smiley, Tuka Alhanai, Ivan Brugere, Armineh Nourbakhsh, Mohammad M. Ghassemi
IEEE TCSS 2025JournalFinance
- Paper
- arXiv
MambaNet: A Hybrid Neural Network for Predicting the NBA Playoffs
Reza Khanmohammadi, Sari Saba-Sadiya, Sina Esfandiarpour, Tuka Alhanai, Mohammad M. Ghassemi
SN Computer Science 2024JournalCore CS
- Paper
The Broad Impact of Feature Imitation: Neural Enhancements Across Financial, Speech, and Physiological Domains
Reza Khanmohammadi, Tuka Alhanai, Mohammad M. Ghassemi
2023PreprintCore CS
- arXiv
Fetal Biological Sex Identification using Machine and Deep Learning Algorithms on Phonocardiogram Signals
Reza Khanmohammadi, Mitra Sadat Mirshafiee, Mohammad M. Ghassemi, Tuka Alhanai
2021PreprintMedicalPresented at IEEE BHI 2023
- arXiv
- Video

Presentations · 7x

+1 more photo
ASTRO 2026 Annual Meeting
Teaching Language Models Safe Clinical Decision-Making
OralSep 2026Paper
Risk-Controlled Selective Automation of Oncology Chart Review
SpotlightSep 2026Paper
Boston
MSU AI Literacy course
Validating AI Outputs: What a Model's Confidence Can and Cannot Tell Us
Guest lecture2026
Video
+1 more photo
MSU and Henry Ford Health Precision Health Symposium
Enhancing LLM Reliability in Precision Health: Calibrating Confidence by Probing Perturbed Representation Stability
PosterOct 2025
Paper
East Lansing
Photos (2)
IEEE BHI 2025
Hybrid Student-Teacher LLM Refinement for Cancer Toxicity Symptom Extraction
TalkOct 2025Paper
NextGen Scholar talk
TalkOct 2025
Video
Atlanta
ASTRO 2024 Annual Meeting
A Novel Localized Student-Teacher LLM for Enhanced Toxicity Extraction in Radiation Oncology
PosterSep 2024
Paper
Washington DC
Henry Ford + MSU Cancer Research Symposium
Iterative Prompt Refinement for Radiation Oncology Symptom Extraction
Invited talkJan 2023
Video
East Lansing
IEEE BHI 2023
Fetal Biological Sex Identification using Machine and Deep Learning Algorithms on Phonocardiogram Signals
Rapid-fire2023
Video Paper
Pittsburgh