Glossary

A Multi-Agent System (MAS) involves multiple interacting decision-makers, like AI models and robots, striving to achieve common or conflicting goals in a shared environment.

Mt-Bench (Multi-Turn Benchmark) captures the ability of large language models to partake in engaging, informative and coherent multi-turn dialogues.

Mteb or Massive Text Embedding Benchmark is a specialized evaluative benchmark, aiding in assessing the performance of diverse text embedding models across various tasks and datasets.