News

  • NeuroX Toolkit on pip: https://pypi.org/project/neurox/
  • Tutorial @NAACL 2021: Fine-grained Interpretation and Causation Analysis in Deep NLP Models
  • Accepted Papers 2021: 1 TACL, 1 NAACL, 1 ACL findings, 2 ICWSM, 1 IWCS
  • Talk: Exploiting Redundancy in Pre-trained Models for Efficient Transfer Learning. Facebook (Feb. 2021), Machine Learning and Data Analytics Symposium (Mar. 2021), National Research Council, Canada
  • Talk: Hidden Linguistics in Deep NLP Models. Heinrich-Heine Universität Düsseldorf, Germany
  • Accepted @COLING 2020: Two long papers, 1) Dialect Arabic machine translation, 2) Location mention recognition
  • Accepted @EMNLP 2020: Two long papers on interpretabiliy are accepted. Details coming soon!
  • Media Coverage: One billion tokens translated. Gulf times, Qatar tribune and several TV/radio channels such as Qatar radio and Qatar TV
  • Released: First-ever Dialectal Arabic to English Machine Translation System covering a large variety of Arabic dialects
  • Tech Transfer: Machine Translation Technology to KanariAI
  • Accepted @ACL 2020: Similarity Analysis of Contextual Word Representation Models
  • Talk: Interpreting Deep NLP Models, University of Edinburgh, UK (April 2020)
  • Talk: University of Sheffield, UK (Mar 2020)
  • Invited talk: Efficient Transfer Learning of Pretrained Model. 7th International Conference on Language and Technology, Pakistan (February 2020)
  • Talk: Analyzing Individual Neurons in Deep NLP Models at Google, Facebook, Amazon, Salesforce and Bosch, US (April 2019)
  • Keynote: Hidden Linguistic in Deep NLP Models, Symposium on Natural Language Processing, University of Moratuwa, Sri Lanka (March 2019)
  • Deep learning for NLP: Fun teaching at International Spring School in Advanced Language Engineering, University of Moratuwa (March 2019)
  • Talk: Machine Translation in Real World, King’s College London, UK (March 2019)
  • Talk: Analyzing Individual Neurons in Deep NLP Models, University of Melbourne, Australia (February 2019)
  • Accepted @NAACL 2019: “One Size Does Not Fit All: Comparing NMT Representations of Different Granularities”
  • Accepted @NAACL 2019: “Highly Effective Arabic Diacritization using Sequence to Sequence Modeling”
  • Media coverage: Our work on NeuroX - Analyzing and Controlling Individual Neurons is featured at MIT news and several AI blogs.
  • Accepted @ICLR 2019: “Identifying and Controlling Important Neurons in Neural Machine Translation”
  • Accepted @AAAI 2019: “What Is One Grain of Sand in the Desert? Analyzing Individual Neurons in Deep NLP Models”
  • Accepted @AAAI 2019: “NeuroX: A toolkit for Analyzing Individual Neurons in Neural Network”
  • Best Innovation Award @ARC’18 for our Speech Translation System, see media coverage
  • Deep learning for NLP: I will be delivering an intensive course on deep learning for NLP at the University of Duisburg-Essen, Germany in the second week of April 2018
  • Accepted @NAACL 2018: Incremental decoding and training methods for simultaneous translation in neural machine translation
  • Media coverage: Our work on understanding Neural Machine Translation has made it to MIT news and been picked by several channels like ScienceBlog, ScienceDaily
  • Accepted @IWSLT 2017: Neural Machine Translation Training in a Multi-domain Scenario
  • Deep learning for MT: I had great fun in teaching deep learning for machine translation at the DGfS-CL Fall School. Course material is available here.
  • Accepted two papers @ACL 2017
  • Read my post about What does Neural Machine Translation Learn about Morphology
Avatar
Hassan Sajjad
Senior Research Scientist

My research interests include natural language processing, applied machine translation and deep learning