About
I am a recent Computer Science PhD graduate, and currently an Applied Science intern at the Amazon Alexa knowledge group in Cambridge, UK.
During my PhD at the University of Machester, I was supervised by Andre Freitas and Uli Sattler. Before that, I did an MSc in mathematics at the University of Stellenbosch, South Africa, supervised by Ingrid Rewitzky.
My core interest is Natural Language Processing, but more specifically:
- NLU (Natural Language Understanding), including models for natural language inference, reasoning and interaction with knowledge bases.
- Interpretability of neural NLP models, including topics such probing and representation-level interventions.
Name Confusion?
My Russian name “Юлия” has been translated differently on various legal documents: I’m “Yulia Rozanova” on my Russian passport, and “Julia Rozanova” on my South African identity documents. I have used “Julia” all my life, but suddenly had to start using “Yulia” legally when I came to the UK in 2019. Weird! Call me Julia. :)
Curriculum Vitae
My latest CV is available here.
Publications
As a First Author:
Estimating the Causal Effects of Natural Logic Features in Neural NLI Models
Preprint
Interventional Probing in High Dimensions: An NLI Case Study
Findings of EACL 2023
Decomposing Natural Logic Inferences in Neural NLI
BlackBoxNLP @ EMNLP 2022
Does my Representation Capture X? Probe-Ably.
ACL 2021 (System Demo Track)
Supporting Context Monotonicity Abstraction in Neural NLI
NaLOMa Workshop @ IWCS 2021
As an Et Al:
Systematicity, Compositionality and Transitivity of Deep NLP Models: a Metamorphic Testing Perspective
Findings of ACL 2022
To Be or Not To Be and Integer? Encoding Variables for Mathematical Text
Findings of ACL 2022
On the Evaluation of Intelligent Process Automation
Workshop on Intelligent Process Automation @ AAAI 2020
∂-Explainer: Abductive Natural Language Inference via Differentiable Convex Optimization
UNDER REVIEW (ACL 2022), Preprint Available
Dissertations
Updates
- 5 December 2023: I am starting as an Applied Science Intern at the Amazon Alexa knowledge group in Cambridge, UK!
- 30 October 2023: My PhD Degree has been formally awarded!
- 7 June 2023: I successfully defended my PhD thesis, examined by Prof Marco Kuhlmann and Dr Ian Pratt-Hartmann.
- 21 January 2023: My paper Interventional Probing in High Dimensions: an NLI Case Study was accepted to Findings of EACL 2023!
- 8 December 2022: I am presenting Decomposing Natural Logic Inferences in Neural NLI at the BlackBoxNLP workshop at EMNLP 2022 in Abu Dhabi.
- 24 February 2022: Two papers that I have been involved in (Systematicity, Compositionality and Transitivity of Deep NLP Models: a Metamorphic Testing Perspective and To Be or Not to Be and Integer? Encoding Variables for Mathematical Text) have been accepted at Findings of ACL 2022!
- 1 January 2022: I joined the Enncore Project as a part-time Research Associate.
- 3 August 2021: I am presenting Probe-Ably at the system demonstration poster session at ACL 2021
- 16 June 2021: I am presenting “Supporting Context Monotonicity Abstraction” at the NaLoMa workshop at IWCS 2021.
- 6 June 2021: Probe-Ably has been accepted at ACL 2021 (System Demo Track)!
- 1 March 2020: I have started as a part-time Research Associate (RA) under Andre Freitas in collaboration with Blue Prism, working on a project concerning Intelligent Process Automation.
- 1 April 2019: I am starting a PhD under Andre Freitas and Uli Sattler at the University of Manchester, UK!