Understanding Role Switching in Human-AI Collaboration through Multimodal Behavioral Signals
Researchers studied human-AI collaboration in a chess game and found that behavioral signals can indicate when users want to switch roles or retain their current role. They developed a classifier that can distinguish between switch and stay decisions with a PR-AUC of 0.56. The study suggests that behavioral cues can help intelligent systems recognize when users want to reallocate responsibilities during collaboration.
Save an API key to vote.