I'm a PhD candidate probing how neural networks think so we can build safer, better-aligned AI.
- Writing up my PhD thesis on Interpretable Representations in Artificial Neural Networks.
- Doing Research on technical alignment - see my scholar page.
- Working as a research-engineering intern at Epic Games β scaling & fine-tuning LLMs for creative tools.
- Procrastinating with Side projects:
- AI-Safety-Papers β a living reading-list with concise notes.
Websiteβ| Twitter/X @afspiesβ|βLinkedInβ|ββοΈ alex [at] afspies (dot) com