Search papers, labs, and topics across Lattice.
Affiliation:
1
0
3
LLMs consistently underperform in shared-budget reasoning tasks, revealing a critical gap between their single-task capabilities and multi-task resource allocation.
