GPT-4 (2023) › The Solution
History of Language AIAll chapters GPT-4 (2023) › The SolutionOne section · Sections in this chapter3 / 7 ↓
Reading tools Definitions & inline links
Choose your expertise level to adjust how many terms are explained. Beginners see more tooltips, experts see fewer to maintain reading flow. Hover over underlined terms for instant definitions.
Article links
Make inline references clickable
OpenAI disclosed few architectural or dataset details for GPT-4. The public technical report focused on capabilities and predictable scaling, with separate sections on evaluation and safety mitigations.
GPT-4 accepted prompts containing text and images, then generated text responses. Demonstrations included questions about charts, diagrams, and photographs. OpenAI did not publish enough architecture detail to specify how visual representations were connected to the language model.
GPT-4 outperformed GPT-3.5 on many reported exams and benchmarks, including tasks involving mathematics and logical questions. It still made reasoning errors, and the public report did not attribute the gains to a specific architectural change.
The reported post-training process included supervised fine-tuning and RLHF, followed by safety evaluation and mitigation work. These steps changed response behavior but could not guarantee reliability or alignment with every user's values.
The development process targeted behavior often summarized as helpful, harmless, and honest. OpenAI used red-team testing, safety evaluations, post-training, and deployment mitigations to reduce harmful responses. The technical report also documented remaining risks and known failure modes.
OpenAI reported GPT-4 around the 90th percentile on the Uniform Bar Exam and 88th percentile on the LSAT, along with gains on other exams. Percentile comparisons depend on the test version, prompting method, and scoring procedure; they do not measure all skills needed in professional practice.
Image input let GPT-4 answer questions about visual material or combine it with text. The feature broadened the kinds of prompts the model could accept, while retaining familiar failure modes such as incorrect descriptions and unsupported inferences.
These capabilities supported experiments in tutoring, writing assistance, software development, document analysis, and other applications. Each use still required task-specific evaluation and safeguards rather than assuming benchmark gains would transfer directly.