DeepMind introduces the 280-billion-parameter Gopher language model
DeepMind introduced Gopher, a 280-billion-parameter language model, and published research on its capabilities, limitations, and risks without releasing public weights or an API.
What happened
Event details
DeepMind published its Gopher research on December 8, 2021. Gopher was a 280-billion-parameter Transformer language model used to study how scale affected reading comprehension, fact-checking, language modeling, and risk-related behavior. DeepMind published the papers and evaluations but did not provide model weights or a public access endpoint.
Assessment
Why it matters
Gopher offered a systematic study of the capabilities and social risks of very large language models and influenced subsequent research on scaling and safety.
Availability
Access notes
The research was public, but the model itself was not available for public use.