Statements (50)
| Predicate | Object |
|---|---|
| gptkbp:instanceOf |
gptkb:report
|
| gptkbp:author |
gptkb:OpenAI
|
| gptkbp:describes |
gptkb:GPT-4
benchmarks evaluation methods ethical considerations capabilities of GPT-4 limitations of GPT-4 multimodal abilities safety mitigations |
| https://www.w3.org/2000/01/rdf-schema#label |
GPT-4 technical report
|
| gptkbp:language |
English
|
| gptkbp:mentions |
gptkb:ChatGPT
gptkb:GPT-3.5 gptkb:OpenAI_API accessibility robustness training data alignment safety research deployment harmful content prompt engineering human feedback model scaling evaluation datasets factuality limitations in bias limitations in code generation limitations in context length limitations in factual accuracy limitations in hallucination limitations in mathematics limitations in multilingual tasks limitations in reasoning limitations in safety alignment limitations in toxicity limitations in vision tasks model architecture system card |
| gptkbp:publicationDate |
2023-03-27
|
| gptkbp:publisher |
gptkb:OpenAI
|
| gptkbp:subject |
gptkb:artificial_intelligence
gptkb:machine_learning natural language processing large language models |
| gptkbp:title |
gptkb:GPT-4_Technical_Report
|
| gptkbp:url |
https://cdn.openai.com/papers/gpt-4.pdf
|
| gptkbp:bfsParent |
gptkb:Jeffrey_Wu
|
| gptkbp:bfsLayer |
7
|