Search papers, labs, and topics across Lattice.
1
1
3
1
LLMs can learn to generate high-quality symbolic world models by interacting with a multi-agent system that provides adaptive, behavior-aware feedback, closing the gap between static validation and interactive execution.
