Portrait of Pere Martra

Pere Martra

ML Research Engineer — LLM optimization & structural pruning

Author of "Rearchitecting LLMs" (Manning) · Barcelona

I research how to make language models smaller, faster and fairer by intervening on their structure: pruning, distillation and activation analysis.

Research

Talks

Invited speaker, STAC Summit London (Apr 2025) · Speaker, Dev.BG Conference (May 2026)

Books

Open source

Demos & models (Hugging Face)

Writing

Medium

Contact

Open to research collaborations.