publications

2026

  1. ArXiv
    Characterizing Web Search by Conversational LLM Agents: From Search Decisions and Strategies to Results and Responses
    Mahsa Amani, Seungeon Lee, Abhisek Dash, Asmaa El Fraihi, Yunah Jang, Elisabeth Kirsten, Qinyuan Wu, Krishna P. Gummadi, Manish Gupta, Abhilasha Ravichander, and 2 more authors
    ArXiv, 2026
    Under Review
    arXiv Summary Code
  2. ArXiv
    To Call or Not to Call: A Framework to Assess and Optimize LLM Tool Calling
    Qinyuan Wu, Soumi Das, Mahsa Amani, Arijit Nag, Seungeon Lee, Krishna P. Gummadi, Abhilasha Ravichander, and Muhammad Bilal Zafar
    ArXiv, 2026
    Under Review
    arXiv Summary Code
  3. NeurIPS
    Understanding the Interplay between Memorization and Learning in Large Language Models
    Bishwamittra Ghosh, Soumi Das, Qinyuan Wu, Mohammad Aflah Khan, Krishna P. Gummadi, Evimaria Terzi, and Deepak Garg
    The Fortieth Annual Conference on Neural Information Processing Systems Evaluations and Datasets Track, 2026
    Early version in The Impact of Memorization on Trustworthy Foundation Models: ICML 2025 Workshop
    arXiv
  4. ICML
    Temporal Context Reinstatement Drives Episodic-Like Order Memory in Long-Context Language Models
    Mathis Pink, Vy A. Vo, Qinyuan Wu, Jianing Mu, Javier S. Turek, Uri Hasson, Kenneth A. Norman, Sebastian Michelmann, Alexander Huth, and Mariya Toneva
    Forty-third International Conference on Machine Learning (ICML 2026), 2026
    Earlier version: Assessing Episodic Memory in LLMs with Sequence Order Recall Tasks (2024), arXiv:2410.08133
    arXiv HTML Supp Code
  5. ACL
    Fine-tuning vs. In-context Learning in Large Language Models: A Formal Language Learning Perspective
    Bishwamittra Ghosh, Soumi Das, Till Speicher, Qinyuan Wu, Mohammad Aflah Khan, Deepak Garg, Krishna P. Gummadi, and Evimaria Terzi
    [Oral] [SAC highlight] The 64th Annual Meeting of the Association for Computational Linguistics (ACL 2026), 2026
    arXiv HTML Code
  6. ACL
    Characterizing Web Search in The Age of Generative AI
    Elisabeth Kirsten, Jost Große Perdekamp, Qinyuan Wu, Mihir Upadhyay, Krishna P. Gummadi, and Muhammad Bilal Zafar
    The 64th Annual Meeting of the Association for Computational Linguistics, Findings (ACL 2026), 2026
    Early version at Agents in the Wild: Safety, Security, and Beyond workshop in ICLR 2026
    arXiv Code
  7. ICLR
    In Agents We Trust, but Who Do Agents Trust? Latent Source Preferences Steer LLM Generations
    Mohammad Aflah Khan, Mahsa Amani, Soumi Das, Bishwamittra Ghosh, Qinyuan Wu, Krishna P. Gummadi, Manish Gupta, and Abhilasha Ravichander
    The Fourteenth International Conference on Learning Representations, ICLR 2026, April 23-27, 2026, Rio de Janeiro, Brazil, 2026
    Present in The International Association for Safe & Ethical AI second annual conference (IASEAIʼ26), non-archival
    arXiv HTML Code Poster Slides
  8. ICLR
    Rote Learning Considered Useful: Generalizing over Memorized Data in LLMs
    Qinyuan Wu, Soumi Das, Mahsa Amani, Bishwamittra Ghosh, Mohammad Aflah Khan, Krishna P. Gummadi, and Muhammad Bilal Zafar
    The Fourteenth International Conference on Learning Representations, ICLR 2026, April 23-27, 2026, Rio de Janeiro, Brazil, 2026
    Present in The International Association for Safe & Ethical AI second annual conference (IASEAIʼ26), non-archival
    arXiv Supp Video Code Poster Slides
  9. WWW
    Algorithmic Self-Portrait: Deconstructing Memory in ChatGPT
    Abhisek Dash*, Soumi Das*, Elisabeth Kirsten*, Qinyuan Wu*, Sai Keerthana Karnam, Krishna P. Gummadi, Thorsten Holz, Muhammad Bilal Zafar, and Savvas Zannettou
    ACM Web Conference 2026, Dubai, UAE, 2026
    Present in The International Association for Safe & Ethical AI second annual conference (IASEAIʼ26), non-archival
    arXiv Video Code Slides
  10. IASEAI
    Revisiting Privacy, Utility, and Efficiency Trade-offs when Fine-Tuning Large Language Models
    Soumi Das, Camila Kolling, Mohammad Aflah Khan, Mahsa Amani, Bishwamittra Ghosh, Qinyuan Wu, Till Speicher, and Krishna P. Gummadi
    The International Association for Safe & Ethical AI second annual conference (IASEAIʼ26), 2026
    arXiv Video

2025

  1. ArXiv
    Position: Episodic Memory is the Missing Piece for Long-Term LLM Agents
    Mathis Pink, Qinyuan Wu, Vy Ai Vo, Javier Turek, Jianing Mu, Alexander Huth, and Mariya Toneva
    2025
    arXiv
  2. WSDM
    Towards Reliable Latent Knowledge Estimation in LLMs: Zero-Prompt Many-Shot Based Factual Knowledge Extraction
    Qinyuan Wu, Mohammad Aflah Khan, Soumi Das, Vedant Nanda, Bishwamittra Ghosh, Camila Kolling, Till Speicher, Laurent Bindschaedler, Krishna P Gummadi, and Evimaria Terzi
    The 18th ACM International Conference on Web Search and Data Mining (WSDM 2025), Hannover, Germany, 2025
    arXiv HTML Supp Code Poster

2024

  1. ArXiv
    Understanding Memorisation in LLMs: Dynamics, Influencing Factors, and Implications
    Till Speicher, Mohammad Aflah Khan, Qinyuan Wu, Vedant Nanda, Soumi Das, Bishwamittra Ghosh, Krishna P Gummadi, and Evimaria Terzi
    arXiv preprint arXiv:2407.19262, 2024
    arXiv Code

2022

  1. Journal
    Exponential negation of a probability distribution
    Qinyuan Wu, Yong Deng, and Neal Xiong
    Soft Computing, 2022
    HTML