Search papers, labs, and topics across Lattice.
7
0
7
MiniOpt achieves the highest solving accuracy for compact models while requiring significantly fewer training resources than traditional methods.
Interactive dialogue can unlock creative potential that static assessments overlook, leading to richer evaluations of creativity in AI contexts.
LLMs struggle to accurately evaluate real human reasoning in mathematics, revealing a critical evaluation gap that challenges current assessment methods.
LLMs can significantly enhance their collaborative performance, achieving over 24% better engagement with human partners when trained in realistic game scenarios.
Scaling offline MARL to thousands of agents is now tractable: MF-Diffuser uses mean-field theory to plan in trajectory distribution space, sidestepping the curse of dimensionality.
AgentSchool offers a powerful new way to simulate educational environments, moving beyond simple role-play to model learning as a dynamic state transition and providing a testbed for long-horizon memory and multi-agent coordination.
Current warehouse robot coordination algorithms break down when faced with realistic constraints like dynamic order streams and physics-aware motion, revealing a critical gap in existing benchmarks.
