Agency science in LLMs

I am currently looking for collaborators (or students) for several ongoing projects:

​

• Mechanistic interpretability in DeepRL and LLMs

  • Continuing to develop a neuroscience of NNs that is neuroscience inspired. For more see on the technical side see this mechint paper and the agency hackathon.

​

• Agency foundation research on a variety of topics. See here for several topics on game theory, RL/IRL, and conceptual research.

​

​

Funding may be available from AI safety/alignment organizations. Contact me to chat more: cm5635(at)nyu(dot).edu(ignore), or find me on the AI alignment slack channel.