
Armineh Nourbakhsh
Graduate student @ CMU, J.P. Morgan AI Research
quantitative reasoning
question answering
metric learning
counterfactual scenarios
compositional attention
finqa
compositional generalization
2
presentations
3
number of views
SHORT BIO
Armineh Nourbakhsh is a PhD student at the Language Technologies Institute at Carnegie Mellon University, as well as a Director at J.P. Morgan AI Research, where she leads a team on multimodal document AI. Her career spans a decade of research in Natural Language Processing in areas such as targeted sentiment analysis, event detection and verification, information extraction, and social data mining. Prior to J.P. Morgan, Armineh was a Director of Data Science at S&P Global, where she led efforts to transform operational workflows related to the ingestion and processing of financial disclosures. In addition to numerous publications and patents, Armineh’s research has been deployed in award-winning AI-driven technologies such as Reuters Tracer, Westlaw Quick Check, and the SocialZ Verve index. She has previously organized workshops at AAAI and ICAIF, and served on the Program Committee of several conferences including IJCAI, AAAI, and ICAIF.
Presentations

Using counterfactual contrast to improve compositional generalization for multi-step quantitative reasoning
Armineh Nourbakhsh and 2 other authors

Improving compositional generalization for multi-step quantitative reasoning in question answering
Armineh Nourbakhsh and 3 other authors