Search papers, labs, and topics across Lattice.
Peking University
1
0
2
Current LLMs fail to effectively anticipate user needs, with even the latest models only succeeding in 26.7% of proactive scenarios.
