Observed arrival · 2026-09-06
Ciara Rowles Studies the Hidden Machinery of Models
A machine learning research portfolio covering generative models, multimodal systems, and mechanistic interpretability.
Field notes
The portfolio organizes work by concrete model behavior rather than broad commercial outcomes. Listed projects cover conditioning and control in diffusion models, text-to-audio alignment for Foley, and image-layer decomposition guided by vision-language model feedback. Its interpretability work asks how internal states change when a language model is instructed to suppress them, and whether monitors trained in one behavioral regime continue to function in another. Code and experiments are directed to GitHub, with papers collected separately.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
One card from the complete issue