[hangonhn]

How is different and/or better than the LLM benchmark released by Spring Health? Github: https://github.com/SpringCare/VERA-MH

The architecture and evaluation approach seem broadly similar.