Amal Alabdulkarim

Ph.D. Candidate in Computer Science
Georgia Institute of Technology

About

I design methods to explain the decisions of AI systems, and I study when those explanations actually help the people who rely on them.

I’m a Ph.D. candidate in Computer Science at the Georgia Institute of Technology, advised by Professor Mark O. Riedl. My dissertation, Temporality in Sequential Explainable AI, explores how agents’ actions can be explained through past events and future expectations, and how the timing of those explanations shapes collaboration in human-AI teams.

My research brings together natural language processing, explainable AI, and reinforcement learning. Recently, I’ve been studying interactive and counterfactual explanations for LLM-based agentic workflows.

  • Explainable AI
  • Natural Language Processing
  • Reinforcement Learning
  • Human-AI Interaction
  • Large Language Models
  • Story Generation

Research

Temporality in Sequential Explainable AI

Temporally grounded agents explanations that link past events and future expectations to current decisions, paired with empirical study of how explanation timing affects user trust and collaboration in human-AI teams.

Explanations for Agentic AI

An explanation framework for multi-step agentic AI pipelines that produces interactive and counterfactual explanations, aimed at improving human oversight and transparency over the decision chains of LLM agents.

Enhanced Reader Models for Story Understanding and Generation

A causal-temporal knowledge structure for narrative understanding that removes the need for full-context dependency, enabling more efficient coherent story generation and reader-model-guided comprehension.

Publications

2026

  1. Ehsan, U., Alabdulkarim, A., Holstein, K., Lee, M. K., Riener, A., & Weisz, J. D. Human-Centered Explainable AI (HCXAI): Re-examining XAI in the Era of Agentic AI. CHI 2026 Extended Abstracts.
  2. Singh, M., Kim, G. C., Okamoto, M., Ammavajjala, A., Alabdulkarim, A., Mansi, G., & Riedl, M. O. Counterfactual Explanations for Agentic Workflows. CHI 2026 Workshop on Human-Centered Explainable AI (HCXAI).

2025

  1. Alabdulkarim, A., Singh, M., Mansi, G., Hall, K., Ehsan, U., & Riedl, M. O. Experiential Explanations for Reinforcement Learning. Neural Computing and Applications, Springer.
  2. Singh, M., Alabdulkarim, A., Mansi, G., & Riedl, M. O. Explainable Reinforcement Learning Agents Using World Models. IJCAI Workshop on Explainable AI (XAI 2025).

2023

  1. Alabdulkarim, A., Mansi, G., Hall, K., & Riedl, M. O. Experiential Explanations for Reinforcement Learning. IJCAI Workshop on Explainable AI (XAI 2023).

2022

  1. Peng, X., Xie, K., Alabdulkarim, A., Kayam, H., Dani, S., & Riedl, M. O. Guiding Neural Story Generation with Reader Models. Findings of ACL: EMNLP 2022.

2021

  1. Alabdulkarim, A., Li, W., Martin, L. J., & Riedl, M. O. Goal-Directed Story Generation: Augmenting Generative Language Models with Reinforcement Learning. arXiv:2112.08593.
  2. Alabdulkarim, A., Li, S., & Peng, X. Automatic Story Generation: Challenges and Attempts. Narrative Understanding Workshop, ACL 2021.
  3. Alhindi, T., Alabdulkarim, A., Alshehri, A., Abdul-Mageed, M., & Nakov, P. AraStance: A Multi-Country Dataset for Arabic Stance Detection. ACL Workshop NLP4IF 2021.

2019

  1. Alabdulkarim, A., & Alhindi, T. Spider-Jerusalem at SemEval-2019 Task 4: Hyperpartisan News Detection. SemEval 2019, ACL.

Contact

I’d love to hear from you! Whether you have questions about my work, are interested in collaborating or mentorship, or simply want to connect, feel free to get in touch. Your message is always welcome.