Holistic Evaluation of Language Models (HELM), a framework to increase the transparency of language models (https://arxiv.org/abs/2211.09110). This framework is also used to evaluate text-to-image models in Holistic Evaluation of Text-to-Image Models (HEIM) (https://arxiv.org/abs/2311.04287).
This adds models such as google/gemini-1.5-pro-001-safety-default and google/gemini-1.5-pro-001-safety-block-none, which are variants of the base model with different safety settings.
This adds models such as
google/gemini-1.5-pro-001-safety-default
andgoogle/gemini-1.5-pro-001-safety-block-none
, which are variants of the base model with different safety settings.