Location via proxy:   
[Report a bug]   [Manage cookies]                
Anastasiia Filippova

Anastasiia Filippova

I recently completed my Master’s in Data Science at Ecole Polytechnique Fédérale de Lausanne. I was fortunate to interned at Apple Machine Learning Research in California under the leadership of Samy Bengio. I was working with Ronan Collobert and his amazing team on efficient training and inference of Large Language Models (LLMs).

I also was lucky to serve as a research assistant at the Mathis Laboratory of Adaptive Intelligence, led by Prof. Mackenzie Mathis, during my Master’s. My work was focused on self-supervised pre-training on neural and behavioral data in collaboration with Steffen Schneider and on foundation models for animal pose estimation with Shaokai Ye.

Previously, I was a Quantitative Research Intern at WorldQuant, where I specialized in fine-tuning LLMs using proprietary financial data.

I earned my BSc in Applied Mathematics and Physics from the Moscow Institute of Physics and Technology. During my undergraduate studies, I was an AI/ML Researcher at the Machine Intelligence Lab, focusing on training and deploying AI models on wearable devices like phones, fitness watches, and trackers.

My areas of interest include efficient training, fine-tunining and inference of large models, self-supervised learning and identifiable representation learning. I am also passionate about AI4Science, encompassing applications in neuroscience, biology and physics.

I am pasionate about making AI efficient and accesible to everyone.

You can know more about me here: anastasiia.filippovaa@gmail.com.

Interests
  • Efficiency
  • Large scale training
  • Self-supervised learning
  • Identifiable Representation Learning
  • AI4Sciece
Education
  • MSc in Data Science, 2021 - 2024

    Ecole polytechnique fédérale de Lausanne

  • BSc in Applied Math and Physics, 2017 - 2021

    Moscow Institute of Physics and Technology

Publications

(2024). No Need to Talk: Asynchronous Mixture of Language Models.

PDF Custom Link

(2024). SuperAnimal pretrained pose estimation models for behavioral analysis. Nature Communications.

PDF Video Custom Link

(2022). iEval: Interactive Evaluation Framework for Open-Domain Empathetic Chatbots. In Proceedings of the 23rd Annual Meeting of the Special Interest Group on Discourse and Dialogue.

PDF Cite Video Custom Link