Search papers, labs, and topics across Lattice.
1
63
3
4
Forget brittle, hand-engineered tool use: ARTIST uses reinforcement learning to teach LLMs to *autonomously* decide when and how to use external tools, leading to a 22% performance boost.