Hello! I'm Jo, undergrad at UChicago studying math and philosophy. I'm currently doing research to better understand how RL on narrow tasks change LLM's safety-relevant behaviours, fortunately mentored by James Evans and Austin Kozlowski at Knowledge Lab.
I co-started and co-lead the AI Safety Group at UChicago XLab. Always happy to talk about any aspects of the group, and let me know if I can be helpful in any way.
I want your feedback - and I want to make it easy for you. You can share your thoughts anonymously here. I promise to 1) take what you say seriously, 2) not guess who you are, 3) assume the best of intentions.
A list of things in my headspace: AI Safety, meditation, metacognition, Kant, Borges, my time at UWC.
Now your turn: what's 1) the question you woke up to 2) your favourite art piece (here's mine), or 3) a cool pattern you found in Conway's Game (here's one)—