Publications

Publications in reversed chronological order. "*" Indicates equal contributions.

Long Papers

2024

  1. ICLR
    ZipIt! Merging Models from Different Tasks without Training
    George Stoica*, Daniel Bolya*, Jakob Bjorner, Pratik Ramesh, Taylor Hearn, and Judy Hoffman
    ICLR 2024

2021

  1. AAAI
    Re-TACRED: Addressing Shortcomings of the TACRED Dataset
    George Stoica, Emmanouil Antonios Platanios, and Barnabás Póczos
    Proceedings of the AAAI Conference on Artificial Intelligence 2021

2020

  1. arXiv
    Improving Relation Extraction by Leveraging Knowledge Graph Link Prediction
    George Stoica, Emmanouil Antonios Platanios, and Barnabás Póczos
    arXiv 2020
  1. AAAI
    Contextual Parameter Generation for Knowledge Graph Link Prediction
    George Stoica*, Otilia Stretcu*, Emmanouil Antonios Platanios*, Tom Mitchell, and Barnabás Póczos
    Proceedings of the AAAI Conference on Artificial Intelligence 2020

Short Papers

2023

  1. UniReps
    ZipIt!: Multitask Model Merging without Training
    George Stoica*, Daniel Bolya*, Jakob Bjorner, Pratik Ramesh, Taylor Hearn, and Judy Hoffman
    (Oral) NeurIPS Unifying Representations in Neural Models Workshop 2023
  2. 2022

    1. VTTA
      Bi-Directional Self-Attention for Vision Transformers
      George Stoica, Taylor Hearn, Bhavika Devnani, and Judy Hoffman
      (Best Paper Award) NeurIPS Vision Transformers Theory and Applications Workshop 2022
    2. 2020

      1. KR2ML
        Re-TACRED: A New Relation Extraction Dataset
        George Stoica, Emmanouil Antonios Platanios, and Barnabás Póczos
        NeurIPS Knowledge Representation Meets Machine Learning Workshop 2020
      2. KR2ML
        Knowledge Graph Enhanced Relation Extraction
        George Stoica, Emmanouil Antonios Platanios, and Barnabás Póczos
        In NeurIPS Knowledge Representation Meets Machine Learning Workshop 2020

      2019

      1. GRL
        Contextual Parameter Generation for Knowledge Graph Link Prediction
        George Stoica*, Otilia Stretcu*, Emmanouil Antonios Platanios*, Tom Mitchell, and Barnabás Póczos
        NeurIPS Graph Representation Learning Workshop 2019