Amir Gholaminejad

Amir Gholaminejad is a Research Fellow at UC Berkeley, affiliated with the Berkeley Artificial Intelligence Research (BAIR) Lab and Sky Lab, and co-director of the Pallas Lab. He earned his PhD from UT Austin, where his dissertation on large-scale bio-physics-based image segmentation won the university's 2018 Outstanding Dissertation Award. His research spans quantized neural networks, neural ODEs, large-scale agentic systems, and scientific machine learning.

• Melosh Medal Finalist
• Amazon Machine Learning Research Award (2020)
• Best Student Paper Award (SC'17)
• Gold Medal (ACM Student Research Competition)

Research trends focus on Transformer optimization, context length extension, memory-efficient AI, and physics-informed neural networks. He mentors PhD, Masters, and undergraduate students, with alumni now at institutions like Microsoft, Apple, and Google. Recent work includes SqueezeLLM and LLMCompiler for ICML'24, and he teaches AI Systems at Berkeley.