Can LLMs use ToM to genuinely persuade you, or do they just use good rhetoric? In our new preprint, we use the MINDGAMES framework to test this. Surprisingly, LLMs like o3 can be incredibly effective persuaders *without* actually understanding your mental states. 🧵👇
Check out our work testing Planning Theory of Mind in LLMs and humans!
@jaredlcm.bsky.social is talking about this at #CogSci2025 this Friday at 4pm, and at @colmweb.org in Montreal in October. Don't miss him, his presentations are top shelf.
Lucky to be back from a month-long break, with a paper out in @bigdatasoc.bsky.social! How do academics participate in the construction of ‘AI’ as a research field? We interviewed 90 university-based AI researchers in the UK, US, and Aus. Paper here: journals.sagepub.com/doi/10.1177/...
1/4
@kathyreid.au
I'm at Neurips, presenting this work with @glenberman.bsky.social , Ned Cooper and @wesleydeng.bsky.social . Come check out our poster at the EvalEval workshop on Sunday or DM me to chat!
In a paper we’re workshopping this week at #NeurIPS2024, Ned Cooper, @wesleydeng.bsky.social, and Ben Hutchinson and I ask: what is the model of societal impacts reflected in efforts to evaluate GenAI systems?
Paper: arxiv.org/abs/2410.22985
Workshop: evaleval.github.io
1/5
