Search papers, labs, and topics across Lattice.
1
0
3
Surgeons often *overaccept* flawed LLM outputs in complex cardiac scenarios, suggesting that human-AI collaboration failures pose a greater risk than model inaccuracies alone.
