Search papers, labs, and topics across Lattice.
1
0
3
Reasoning in LLMs can backfire in behavioral simulations, turning them into hyper-rational agents that fail to capture the messy compromises of real-world negotiations.