Big Bench

Study / ResearchMentioned in 1 video

A benchmark with many diverse tasks, some of which are very abstract and unrealistic, which IMB generally avoids for core model evaluation.

Videos Mentioning Big Bench

State of the Art: Training 70B LLMs on 10,000 H100 clusters

State of the Art: Training 70B LLMs on 10,000 H100 clusters

Latent Space

A benchmark with many diverse tasks, some of which are very abstract and unrealistic, which IMB generally avoids for core model evaluation.