publications

publications by categories in reversed chronological order. generated by jekyll-scholar.

2026

  1. Delulu-FIM: A Verified Multilingual Benchmark for Code Hallucination Detection in Fill-in-the-Middle Tasks
    Mahdi Erfanian, Aayush Garg, Xiao Liu, and 2 more authors
    arXiv preprint arXiv:2605.07024, 2026
    Submitted to EMNLP 2026
  2. Synthetic Hallucinations, Real Gains: Hard Negatives from Frontier Models for FIM Hallucination Mitigation
    Mahdi Erfanian, Aayush Garg, Xiao Liu, and 2 more authors
    2026
    Submitted to TMLR
  3. Generative AI for Multi-Modal Data Management
    Mahdi Erfanian
    In VLDB 2026 Ph.D. Workshop, 2026
    Accepted for oral presentation
  4. NeedleDB: A Generative-AI-Based System for Accurate and Efficient Image Retrieval using Complex Natural Language Queries
    Mahdi Erfanian, and Abolfazl Asudeh
    2026
    Preprint

2025

  1. Task-Aware Data Augmentation using Generative AI for Group-Distributional Robustness
    Mahdi Erfanian, Boris Glavic, and Abolfazl Asudeh
    2025
    Submitted to The VLDB Journal
  2. Needle: A Generative-AI-Powered Multimodal Database for Answering Complex Natural Language Queries
    Mahdi Erfanian, Mohsen Dehghankar, and Abolfazl Asudeh
    arXiv preprint arXiv:2412.00639, 2025
    Submitted to TMLR
  3. An Efficient Matrix Multiplication Algorithm for Accelerating Inference in Binary and Ternary Neural Networks
    Mohsen Dehghankar, Mahdi Erfanian, and Abolfazl Asudeh
    In The 2025 International Conference on Machine Learning, 2025
    arXiv preprint arXiv:2411.06360

2024

  1. Chameleon: Foundation Models for Fairness-Aware Multi-Modal Data Augmentation to Enhance Coverage of Minorities
    Mahdi Erfanian, H. V. Jagadish, and Abolfazl Asudeh
    Proceedings of the VLDB Endowment, 2024
    DOI
  2. FairEM360: A Suite for Responsible Entity Matching
    Nima Shahbazi, Mahdi Erfanian, Abolfazl Asudeh, and 2 more authors
    Proceedings of the VLDB Endowment, 2024
    DOI
  3. Coverage-based Data-centric Approaches for Responsible and Trustworthy AI
    Nima Shahbazi, Mahdi Erfanian, and Abolfazl Asudeh
    Bulletin of the IEEE Computer Society Technical Committee on Data Engineering, 2024

2023

  1. Graph Pattern Mining Paradigms: Consolidation and Renewed Bearing
    Vinicius Dias, Samuel Ferraz, Aditya Vadlamani, and 5 more authors
    In 2023 IEEE 30th International Conference on High Performance Computing, Data, and Analytics (HiPC), 2023
    DOI