GPT-4
E19435
artificial intelligence model
generative pre-trained transformer
large language model
multimodal language model
GPT-4 is a large multimodal language model known for its advanced reasoning, comprehension, and generation capabilities across text and images.
Aliases (2)
Statements (48)
| Predicate | Object |
|---|---|
| instanceOf |
artificial intelligence model
→
generative pre-trained transformer → large language model → multimodal language model → |
| accessModel |
cloud-based
→
|
| announcedBy |
OpenAI
→
|
| architecture |
Transformer
→
|
| capability |
code generation
→
conversation → image understanding → natural language generation → natural language understanding → question answering → reasoning → summarization → translation → |
| creatorOrganization |
OpenAI, L.L.C.
→
|
| creatorOrganizationType |
AI research lab
→
|
| designedFor |
general-purpose AI assistance
→
|
| developer |
OpenAI
→
|
| inputType |
image inputs
→
text prompts → |
| license |
proprietary
→
|
| modality |
images
→
text → |
| modelFamily |
GPT series
→
|
| notableProperty |
improved reasoning over GPT-3.5
→
larger context handling than GPT-3.5 → multimodal input support → |
| outputType |
text
→
|
| predecessor |
GPT-3.5
→
|
| provider |
OpenAI API
→
|
| releaseDate |
2023-03-14
→
|
| safetyFeature |
content filtering
→
harm reduction mechanisms → |
| supports |
multiple languages
→
|
| trainingDataType |
articles
→
books → code → web text → |
| trainingMethod |
reinforcement learning from human feedback
→
supervised fine-tuning → unsupervised pre-training → |
| useCase |
chatbots
→
content creation → education assistance → programming assistance → research assistance → |