Hi, my name is Anastasia!
I research how to improve machine’s ability to understand and engage with the world through language. I’m also interested in how our society will be reshaped by having access to and interacting with these increasingly capable AI models.
When GPT-3.5 launched, I shrugged it off as a fun gimmick. Now? I can’t imagine my day without LLMs. These models are rewriting what’s possible in how we interact with technology, approach problem-solving, and even think about creativity. Yet, there are lots of challenges and unanswered questions about their behavior. Here’s what I’m diving into:
Refined control Task performance depends heavily on how prompts are structured. Post-training has advanced models’ instruction-following capabilities, but language is more a bridge than a perfect mirror of thought: some concepts defy precise expression through text, and crafting the right contexts to elicit specific behaviors can be challenging. Furthermore, models remain fragile: small prompt tweaks or adversarial inputs can lead to harmful or erroneous outputs. Research shows that models encode high-level representations of the world. I explore how we can leverage these representations to obtain an additional lever of control on the model’s behavior.
Continual learning Embedding large language models into larger systems with long-term memory and tools for interacting with the world could unlock even more sophisticated task-solving capabilities. Extending context windows is one approach, but it has limits: no window can encompass all past events, nor can we guarantee effective retrieval of events through attention mechanisms. Equally important is the ability to break down complex tasks into structured plans and learn effectively from observations. Enhancing these capabilities is a key area of my research.
I’m an honorary lecturer at Imperial College London, leading a research group tackling these challenges. I also consult on AI deployment, fine-tuning, interpretability, and safety. Previously, I was a senior machine learning scientist at Liquid AI and co-founded Skialabs, where we developed software that manages waste collection operations in cities across the Netherlands, including Amsterdam.
Blog
What began with a younger me writing stories about a mom and dad gifting their daughter a cat (an attempt at manifestation?), has evolved into reflections on machine learning and, occasionally, the quirks of life. Visit my Substack if you’re interested to read more.
Youtube
Prefer to learn about machine learning through videos? Head over to my YouTube channel for deepdives and tutorials on the latest in ML.
Publications
See Google Scholar for a full overview of my more formal research publications.
Get in touch
- Email: a.surname@imperial.ac.uk
I’d love to hear from you! Whether you have an idea to discuss, a problem to brainstorm, or a project to collaborate on, feel free to reach out. To ensure I read your message, include ‘Building better ML’ in the subject line.
Links
This is a jekyll based resume template made by Ankit Sultana