Search papers, labs, and topics across Lattice.
1
0
Agents struggle with long-horizon tasks, achieving only a 15.2% success rate even with advanced models, highlighting a critical gap in current AI capabilities.
