Gopher logo

Gopher

Free

Gopher by DeepMind is a 280 billion parameter language model.

FreeFree tier
Inputs: textOutputs: text
Type
Open Source
Company
Google DeepMind

About Gopher

Gopher is a 280 billion parameter transformer language model developed by DeepMind, released alongside three papers covering language modelling at scale, ethical and social risks, and a new architecture. It demonstrates strong performance on reading comprehension, fact-checking, toxic language identification, and the Massive Multitask Language Understanding (MMLU) benchmark, surpassing prior language models. Gopher also exhibits surprising coherence in dialogue interactions, even without dialogue-specific fine-tuning. However, the model has documented failure modes including repetition, stereotypical biases, and confident propagation of incorrect information. Scale does not significantly improve performance on logical reasoning and common-sense tasks. The research emphasizes the need for interdisciplinary collaboration to anticipate and mitigate risks associated with large language models.

Key Features

280 billion parameter transformer architecture
Strong performance on reading comprehension, fact-checking, and toxic language identification
Significant advancement on Massive Multitask Language Understanding (MMLU) benchmark
Surprising coherence in dialogue interactions
Evaluated for failure modes: repetition, biases, and incorrect information propagation
Part of a comprehensive study on ethical and social risks of large language models

Pros & Cons

Pros
  • Achieves state-of-the-art results on several benchmarks including MMLU
  • Shows strong reading comprehension and fact-checking capabilities
  • Demonstrates coherent dialogue without specific fine-tuning
  • Research includes thorough documentation of failure modes and ethical risks
Cons
  • Tendency toward repetition in outputs
  • Reflects stereotypical biases present in training data
  • Confidently propagates incorrect information
  • Scale does not significantly improve logical reasoning and common-sense tasks

Best For

Summarising informationProviding expert adviceFollowing instructions via natural languageDialogue interaction and chatLanguage modelling research and analysis of scale effects

FAQ

What is Gopher?
Gopher is a 280 billion parameter transformer language model developed by DeepMind, released as part of a research portfolio on language modelling at scale.
What tasks does Gopher perform well on?
Gopher excels at reading comprehension, fact-checking, toxic language identification, and the Massive Multitask Language Understanding (MMLU) benchmark.
What are the limitations of Gopher?
Gopher has failure modes including repetition, stereotypical biases, and confident propagation of incorrect information. It does not show significant improvement on logical reasoning and common-sense tasks.
Was Gopher released as open source?
The website does not specify an open-source release; only the associated papers were released.