Search papers, labs, and topics across Lattice.
IT University of Copenhagen, Denmark
4
0
6
Agentic collectives of LLMs reveal complex behaviors that can be interrogated through their own language, transforming our approach to understanding AI dynamics.
Task-dependent credit assignment reveals hidden bottlenecks in neural network performance, challenging the notion of universal learning strategies.
Modularity in neural architectures can dramatically improve compositional learning, but only in lower-dimensional settings where task representations are rich and nuanced.
LLMs trained with Vector Policy Optimization (VPO) learn to produce diverse solutions that unlock previously unsolvable problems in evolutionary search, outperforming models optimized for single scalar rewards.
