Illia Polosukhin is a software engineer and entrepreneur who was one of the eight co-authors of the seminal 2017 paper "Attention Is All You Need" that introduced the Transformer architecture. His contributions at Google Brain helped shape the implementation and experimental validation of the architecture that would transform natural language processing and artificial intelligence.
Early Life and Education
Born in Ukraine, Polosukhin studied computer science at the Kharkiv National University of Radioelectronics. He moved to the United States and joined Google, where he worked on machine learning infrastructure and neural machine translation as part of the team that developed the Transformer.
Joined Google, working on machine learning and NLP
Co-authored "Attention Is All You Need"
Left Google to co-found NEAR Protocol
NEAR Protocol mainnet launched
Key Contributions
As a co-author of the Transformer paper, Polosukhin contributed to the development and implementation of multi-head attention, the key mechanism that allows the model to jointly attend to information from different representation subspaces. The multi-head attention mechanism computes multiple attention functions in parallel: MultiHead(Q,K,V) = Concat(head_1,...,head_h)W^O, where each head_i = Attention(QW_i^Q, KW_i^K, VW_i^V).
Polosukhin's engineering expertise was critical in implementing the Transformer efficiently on modern hardware and demonstrating its practical effectiveness. The paper showed that Transformers could achieve state-of-the-art translation quality while training significantly faster than recurrent models, a result that depended on both architectural innovation and careful engineering.
"The Transformer was not just a theoretical advance — it was designed from the start to be practically trainable on modern parallel hardware." — On the engineering principles behind the Transformer
Legacy
Polosukhin's contribution to the Transformer paper helped launch the most significant architectural paradigm shift in the history of NLP. After leaving Google, he co-founded NEAR Protocol, applying his machine learning and systems engineering expertise to blockchain technology. His career trajectory illustrates how the skills developed in NLP research transfer to other domains of technology innovation.