André is motivated to understand the representations and computations underpinning brain-AI (mis)alignment using techniques from mechanistic interpretability, especially in the visual domain. Such techniques are adopted to increase the monosemanticity of representations for a more thorough interpretation of this alignment. Also by performing “artificial neuroscience” on AI models, he hopes to discover mechanisms within them that generate hypotheses for how information processing occurs in the brain. In his free time, André enjoys music (especially jazz), books (especially science nonfiction), and biking around town.