Research

Long-horizon tasks and LLM agents.

I’m a Ph.D. student at Arizona State University, advised by Chitta Baral (ASU) and Eduardo Blanco (University of Arizona).

My research focuses on long-term memory and long-horizon agents. I study how a model decides what to keep in memory and what to let go of, and I build tool-using agents that stay reliable and follow their instructions across a long sequence of steps, rather than a single answer. This builds on my earlier work on reasoning over time and understanding events: extracting structured information about events from text, and testing whether language models can really reason about dates, durations, and the order things happen in.

Alongside my PhD, I’ve interned at Microsoft, Cohere, and Genies.

Publications

• From Recall to Forgetting: Benchmarking Long-Term Memory for Personalized Agents
  Md Nayem Uddin, Kumar Shubham, Eduardo Blanco, Chitta Baral, Gengyu Wang
  Findings of the Association for Computational Linguistics (ACL Findings), 2026.
  ACL PDF
• GuidedSampling: Steering LLMs Towards Diverse Candidate Solutions at Inference-Time
  Divij Handa, Mihir Parmar, Aswin RRV, Md Nayem Uddin, Hamid Palangi, Chitta Baral
  International Conference on Learning Representations (ICLR), 2026.
  ICLR PDF
• UnSeenTimeQA: Time-Sensitive Question-Answering Beyond LLMs' Memorization
  Md Nayem Uddin, Amir Saeidi, Divij Handa, Agastya Seth, Tran Cao Son, Eduardo Blanco, Steven Corman, Chitta Baral
  Annual Meeting of the Association for Computational Linguistics (ACL), 2025.
  ACL PDF Code
• BEMEAE: Moving Beyond Exact Span Match for Event Argument Extraction
  Enfa Rose George, Md Nayem Uddin, Oghenevovwe Ikumariegbe, Daniyal Kashif, Eduardo Blanco, Steven Corman
  North American Chapter of the Association for Computational Linguistics (NAACL), 2025.
  NAACL PDF
• ThinkTuning: Instilling Cognitive Reflections without Distillation
  Aswin RRV, Jacob Dineen, Divij Handa, Md Nayem Uddin, Mihir Parmar, Chitta Baral, Ben Zhou
  Conference on Empirical Methods in Natural Language Processing (EMNLP), 2025.
  EMNLP PDF
• When "Competency" in Reasoning Opens the Door to Vulnerability: Jailbreaking LLMs via Novel Ciphers
  Divij Handa, Zehua Zhang, Amir Saeidi, Shrinidhi Kumbhar, Md Nayem Uddin, Aswin RRV, Chitta Baral
  NeurIPS 2025 Workshop on Reliable ML from Unreliable Data.
  NeurIPS-W PDF
• Insights into Alignment: Evaluating DPO and its Variants Across Multiple Tasks
  Amir Saeidi, Shivanshu Verma, Md Nayem Uddin, Chitta Baral
  ACL Student Research Workshop (ACL SRW), 2025.
  ACL-SRW PDF
• Chaos with Keywords: Exposing Large Language Models Sycophancy to Misleading Keywords and Evaluating Defense Strategies
  Aswin RRV*, Nemika Tyagi*, Md Nayem Uddin*, Neeraj Varshney, Chitta Baral (*equal contribution)
  Findings of the Association for Computational Linguistics (ACL Findings), 2024.
  ACL-F PDF
• Generating Uncontextualized and Contextualized Questions for Document-Level Event Argument Extraction
  Md Nayem Uddin, Enfa Rose George, Eduardo Blanco, Steven Corman
  North American Chapter of the Association for Computational Linguistics (NAACL), 2024.
  NAACL PDF
• Asking and Answering Questions to Extract Event-Argument Structures
  Md Nayem Uddin, Enfa Rose George, Eduardo Blanco, Steven Corman
  Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING), 2024.
  LREC-COLING PDF
• Interpreting Indirect Answers to Yes-No Questions in Multiple Languages
  Zijie Wang, Md Mosharaf Hossain, Shivam Mathur, Terry Cruz Melo, Kadir Bulut Ozler, Keun Hee Park, Jacob Quintero, MohammadHossein Rezaei, Shreya Nupur Shakya, Md Nayem Uddin, Eduardo Blanco
  Findings of the Association for Computational Linguistics: EMNLP, 2023.
  EMNLP-F PDF

Experience

• 
  Microsoft · Applied Scientist Intern

  May 2026 – Present · Redmond, WA

  • Building LLM agents for large-scale system monitoring and automated root-cause analysis across distributed services.
• 
  Cohere · Intern of the Technical Staff

  Jan 2026 – Apr 2026 · San Francisco, CA

  • Implemented continuous learning and performance evaluation for workplace automation agents.
  • Designed long-term memory systems that let tool-calling agents learn from prior experience.
• 
  Genies · Research Scientist Intern

  Jun 2025 – Dec 2025 · San Mateo, CA

  • Built a systematic evaluation framework for LLM-based memory modules (work accepted to ACL 2026).
  • Developed a long-term memory management pipeline for AI agents and user personalization.
• 
  Center for Strategic Communication, ASU · Graduate Research Assistant

  Aug 2022 – Jun 2025 · Tempe, AZ

  • Built a question-answering system to extract event arguments from news articles, with a focus on weak supervision.
  • Led the research behind papers at LREC-COLING 2024, NAACL 2024, NAACL 2025, and ACL 2025.
• 
  Nascenia · Machine Learning Software Engineer

  Jan 2020 – Jun 2022 · Dhaka, Bangladesh

  • Built five conversational agents (the National Digital Service Assistant) for government stakeholders.
  • Set up a data-acquisition web app for end-to-end model training in PyTorch.
  • Created five domain-specific intent-understanding datasets for Bangla.

News

• May 2026Joining Microsoft (Redmond) as an Applied Scientist Intern, working on LLM agents for large-scale system monitoring and automated root-cause analysis.
• Apr 2026"From Recall to Forgetting" (the Memora benchmark) accepted to ACL 2026 (Findings).
• Jan 2026"GuidedSampling" accepted to ICLR 2026.
• Jan 2026Joining Cohere (San Francisco) as an Intern of the Technical Staff.
• Sep 2025"ThinkTuning" accepted to EMNLP 2025.

More news →