Statements (31)
| Predicate | Object |
|---|---|
| gptkbp:instanceOf |
gptkb:large_language_model
|
| gptkbp:basedOn |
gptkb:GPT-4_architecture
|
| gptkbp:developedBy |
gptkb:OpenAI
|
| gptkbp:feature |
scalability
sparse activation expert routing |
| gptkbp:hasModel |
gptkb:transformation
|
| gptkbp:language |
English
multilingual |
| gptkbp:notableFor |
improved performance
efficiency improvements scaling to larger models |
| gptkbp:parameter |
undisclosed
|
| gptkbp:purpose |
gptkb:translator
natural language processing code generation question answering summarization text generation |
| gptkbp:relatedTo |
gptkb:GPT-3
gptkb:GPT-4 Mixture of Experts models |
| gptkbp:releaseYear |
2023
|
| gptkbp:trainer |
gptkb:law
gptkb:Wikipedia books web text |
| gptkbp:uses |
gptkb:Mixture_of_Experts_(MoE)
|
| gptkbp:bfsParent |
gptkb:Mixture_of_experts
|
| gptkbp:bfsLayer |
8
|
| https://www.w3.org/2000/01/rdf-schema#label |
GPT-4 MoE variant
|