Search papers, labs, and topics across Lattice.
3
0
6
3
Iteratively prompting LLMs can either collapse diversity or maintain novelty, revealing a sensitivity to temperature and initial conditions that has implications for multi-agent systems.
LLMs struggle with native Greek, as evidenced by substantial performance gaps revealed by the new GreekMMLU benchmark, highlighting the need for better adaptation and evaluation methods.
MixtureKit offers a unified, open-source toolkit that simplifies the creation and analysis of Mixture-of-Experts models, enabling researchers to easily experiment with different routing strategies and visualize expert behavior.
