I am a Ph.D. researcher in IEMS at the University of Central Florida's Complex Adaptive Systems Lab, advised by Dr. Niloofar Yousefi. My research is about reliable LLMs when humans can't supervise at scale. Four concrete questions drive the work. (1) Can pretrained transformers report calibrated uncertainty in their own attention? (2) Can we verify AI-generated supervision when the judges themselves share blind spots? (3) Can we monitor tool-using agents and intervene on unsafe trajectories, not just outputs? (4) Can we steer a model away from outdated parametric memory at test time, without retraining or retrieval? My applied vertical is clinical and biomedical AI, where the cost of confident errors is measured in patient harm, not benchmarks.

  1. Test-Time Uncertainty in Pretrained Transformers
    • Inference-time uncertainty-aware attention for pretrained transformers UAT-LITE
    • Bayesian uncertainty quantification for safe clinical decision support MedBayes-Lite
  2. Reliable AI Supervision
    • Fault-tolerant preference alignment via multi-agent verification MAV
    • Reliability-aware filtering when LLM judges share blind spots CorrFilter
  3. Runtime Safety and Evaluation for LLM Agents
    • Structured-plan intervention policies for tool-using agents NEXUS
    • Process-level evaluation of clinical LLM agents under uncertainty CareBench
  4. Test-Time Steering Without Retraining
    • Activation steering for parametric temporal conflict in LLMs TAS
    • Closed-form post-training steering when policies cannot be retrained PoE-Steer
📢
Graduating Spring 2028. Seeking a Research Scientist role or Postdoctoral position in LLM training, reinforcement learning, and LLM safety. Please reach out if you have an opening.

Recent updates

Apr 2026
🎉 First-author paper Fault-Tolerant Preference Alignment accepted at ICLR 2026 Trustworthy AI Workshop: multi-agent verification that filters corrupted supervision before RLHF/DPO.
Mar 2026
✨ Two PhD papers out the same month: AMR computational paradigms survey in Briefings in Bioinformatics, and the follow-up agentic RNA-seq paper in Frontiers in Bioinformatics.
Jan 2026
🎤 At Amazon JFK 27 in New York for the Trusted AI Symposium, presenting SAVe V.1: multi-agent safe alignment verification for preference-based LLM optimization.
Oct 2025
🎓 Presented CITE V.1 (evidence-grounded LLM interpretation of RNA-seq clusters) at MIT Jameel Clinic's 7th Molecular Machine Learning Conference.

Featured Publications

Check out Google Scholar for a full list of my publications.

Under Review EMNLP · 2026

UAT-LITE: Inference-Time Uncertainty-Aware Attention for Pretrained Transformers

E. Hossain, S. R. Dipta, S. Neupane, R. Rana, R. Shwartz-Ziv, I. Garibay
ACL ARR meta-review: borderline conference accept · committing to EMNLP 2026

Inference-time method that uses Monte-Carlo dropout to make pretrained transformer attention uncertainty-aware. Improves calibration, selective prediction, and robustness without retraining.

Published BiB · 2026

Computational Paradigms for Antimicrobial Resistance Prediction: Integrating Multi-Omics, Structural Modeling, and Foundation AI Systems

E. Hossain, N. Yousefi
Briefings in Bioinformatics, 2026

Survey unifying multi-omics, structural modeling, and foundation models for AMR prediction; positions where each paradigm shines and where they break.

Published Frontiers · 2026

BIOGEN: Evidence-Grounded Multi-Agent Reasoning Framework for Transcriptomic Interpretation in Antimicrobial Resistance

E. Hossain, M. Shoeibi, I. Garibay, N. Yousefi
Frontiers in Bioinformatics, vol. 6, 2026 · Drug Discovery in Bioinformatics

A multi-agent framework that interprets RNA-Seq gene clusters by combining biomedical retrieval with evidence verification; every cluster interpretation cites at least one PMID, DOI, or UniProt accession across five bacterial datasets.

Preprint NeurIPS · 2026

When Policies Cannot Be Retrained: A Unified Closed-Form View of Post-Training Steering in Offline Reinforcement Learning

E. Hossain, M. J. I. Basher, I. Garibay, O. Garibay, N. Yousefi
Preprint, 2026 · Under review at NeurIPS

A closed-form analysis of post-training policy steering in offline RL, showing when interventions help, hurt, or leave a frozen policy unchanged.

Published Nature SR · 2026

R-GAT: Cancer Document Classification Leveraging Graph-Based Residual Network for Scenarios with Limited Data

E. Hossain, T. Nuzhat, S. Masum, S. Rahimi, N. A. Golilarz
Scientific Reports (Nature), 2026

Residual graph attention network for biomedical document classification under limited data; matches transformer baselines on cancer abstracts at a fraction of the cost.

Published IEEE FIE · 2024

From Questions to Insightful Answers: Building an Informed Chatbot for University Resources

S. Neupane, E. Hossain, J. Keith, H. Tripathi, F. Ghiasi, N. A. Golilarz, A. Amirlatifi, S. Mittal, S. Rahimi
IEEE Frontiers in Education Conference (FIE), 2024

Retrieval-augmented, domain-informed chatbot framework for university support; demonstrates effective conversational agents on grounded institutional knowledge.

Certifications

2023
AI for Medicine Specialization — DeepLearning.AI on Coursera. Three-course track covering medical diagnosis, prognosis, and treatment-effectiveness prediction with deep learning. Verify
2022
IBM Professional Data Science Certificate — Coursera. End-to-end data science workflows: statistical modeling, machine learning with Python, and production pipelines. Verify

Projects

🛡️
2026

SAVe V.1

Multi-agent safe alignment verification for preference-based LLM optimization. Presented as a poster at the Amazon Trusted AI Symposium, JFK 27, New York.

🧠
2025

MiniHealthLM

Lightweight, secure transformer language model trained from scratch for healthcare text. GitHub

📊
2025

mldriguard

Python package for detecting and mitigating data drift in machine learning models. GitHub

💊
2025

MedXpert

AI-powered drug discovery and comparison system. GitHub

Awards / Recognition

2025
Graduate Presentation Fellowship, UCF. Selective award funding research presentation at MIT.
2019
Finalist, Dr. Pradeep P. Thevannoor Innovation Awards, India.
2023
Award, Fourth Industrial Revolution Skills Summit, North South University (NSU).
2018
Winner, IEEE SS12 Innovation Challenge & Maker Fair, for the Kidnap Prevention Mobile App.
2017
Divisional Champion & Global Nominee, NASA Space Apps Challenge, for the Drone for Green Project.