Machiko Hirota
MEng Computer and Information Science, University of Pennsylvania·0 publications
About
Hi, I'm Machiko. I'm a software engineer at Relativity Space and a Technical Governance Research Fellow at MIRI. My research spans AI safety, NLP, and alignment — including work on AI-exacerbated infrastructure vulnerability through the SPAR fellowship. At Shiba AI, I contribute to the mechanistic interpretability, technical governance, and bias/persona groups. Previously, I developed ReazonSpeech-k2-v2 at the Reazon Human Interaction Lab in Tokyo, a large-scale Japanese ASR model, and worked on GPU-accelerated computer vision algorithms at NVIDIA. I also hold an MEng in CIS from UPenn and have done NLP research at the University of Tokyo and Penn's Computational Social Listening Lab.