Search papers, labs, and topics across Lattice.
New York University
1
0
2
LLM agents struggle to uncover hidden environments, with performance plummeting as task complexity increases, revealing fundamental limitations in their interactive reasoning capabilities.
