We show that synthetic data (e.g., LLM simulations) can significantly improve the performance of inference tasks. The key intuition lies in the interactions between the moment residuals of synthetic data and those of real data
We show that synthetic data (e.g., LLM simulations) can significantly improve the performance of inference tasks. The key intuition lies in the interactions between the moment residuals of synthetic data and those of real data
vhil.stanford.edu/publications...
vhil.stanford.edu/publications...
@nathumbehav.nature.com
vhil.stanford.edu/publications...
@nathumbehav.nature.com
vhil.stanford.edu/publications...
Monique Santoso coded 9,000 speech acts to develop Virtual Reality Interaction Dynamics Scheme, 10 speech acts (e.g., disagreements, context-dependent commentary). Prior speech acts and current nonverbal behavior predict group action.
vhil.stanford.edu/publications...
Monique Santoso coded 9,000 speech acts to develop Virtual Reality Interaction Dynamics Scheme, 10 speech acts (e.g., disagreements, context-dependent commentary). Prior speech acts and current nonverbal behavior predict group action.
vhil.stanford.edu/publications...
@PortiaWang.bsky.social analyzed a VR dataset of 77 sessions, 1660 minutes of group meetings over 4 weeks. Verbal & nonverbal history captured at millisecond level predicted turn-taking at nearly 30% over chance. To appear @acm-cscw.bsky.social.
vhil.stanford.edu/publications...
Go Gators! @ufjschool.bsky.social
Go Gators! @ufjschool.bsky.social
vhil.stanford.edu/publications...
vhil.stanford.edu/publications...