OpenAI has recently launched its much-awaited language model, GPT-4, on ChatGPT Plus and API (Waitlist).
People have been talking about its ability to score high on several examinations such as the Uniform Bar exam, SATs, GRE, and more. The AI language model was able to pass the Bar Exam and SAT with the 90th percentile, which is quite impressive.
Additionally, it was able to perform various tests from high school to college level to showcase its capabilities.
In this article, we will explore the examinations performed by GPT-4 and how the multimodal language model performed.
GPT-4 – What, like law school, is hard?
OpenAI’s new language model GPT-4 was easily able to pass the LSATs (Law School Admission Test) by scoring 168 and an estimated percentile of 88th on the test.
Apart from this, GPT-4 was able to score 298/400, with an estimated percentile of 90th in the Uniform Bar Exam (MBE+MEE+MPT).
Previously, OpenAI’s language model, GPT-3.5, scored only 213/400 in the Uniform Bar Exam with an estimated percentile of 10th and 149 in LSAT with a percentile of 40th. This shows the significant improvement in GPT-4’s performance.
College Admissions Tests: A Piece of Cake for GPT-4
College admission tests were taken by GPT-4 to showcase its capabilities in solving complex queries.
The multimodal language model took SATs in math and reading/writing, including three sections of the GRE (Graduate Record Examination).
In the GRE Quantitative test, GPT-4 was able to score a percentile of 88th and 99th on the GRE’s verbal examination.
In addition to the college admission tests, GPT-4 also took tests for high school examinations, including all the AP (Advanced Placement) tests such as Biology, Chemistry, English literature and composition, Calculus BC, Art History, Psychology, and more.
GPT-4 aced AP Biology and Art History by scoring a percentile of 86th -100th in AP Art history and 84th-100th percentile in AP Art History.
In AP Calculus BC, GPT-4 scored a percentile of 43rd-59th. GPT-4 scored 44th percentile in AP English Language and 14th-44th in English literature and composition.
Although GPT-4 wasn’t able to perform amazingly in all the high school examinations, it still performed decently in most of them by scoring a percentile of 86th-100th, which is quite impressive.
GPT-4 Has Some Coding Work to Do
As the field of artificial intelligence continues to progress, we are witnessing new breakthroughs in the development of language models.
OpenAI’s latest innovation, GPT-4, is one such model that has created a buzz in the tech world. With its advanced capabilities in processing visual inputs and generating human-like text, GPT-4 has set new benchmarks in professional and academic examinations. Let’s dive in and explore the capabilities of GPT-4.
Coding Abilities of GPT-4
Although GPT-4 has made considerable progress in language modeling, it still has some coding work to do.
As per the reports, GPT-4 was able to solve 31 out of 41 problems in the easy levels of Leetcode. However, it struggled with medium and hard levels, solving only 21 out of 80 and 3 out of 45 problems, respectively.
With a rating of 392 on Codeforces, which categorizes it in the Newbie level of anything lower than 1199, GPT-4 needs to work on its coding abilities to reach higher levels.
Despite its limitations in coding, GPT-4 showcased its capabilities in writing python during a live stream demonstration for developers.
However, it still requires manual tweaking to generate proper parameters, which could be the reason behind its lower scores in coding. With further development and refinement, GPT-4 could become a top-notch coding assistant.
Multimodal Capabilities of GPT-4
The major factor that distinguishes GPT-4 from other language models is its ability to analyze and understand visual inputs. GPT-4 can process visual inputs like images and can process up to 25,000 words, making it possible to read, summarize, translate, and generate text answers in a human-like way.
This allows GPT-4 to exhibit “human-level performances” in passing exams like bar exams, SATs, and more.
Imagine providing an image of flour and eggs, and GPT-4 generating a recipe for chocolate cakes.
With its ability to understand vision input, GPT-4 can comprehend questions based on images and provide a step-by-step answer in a human-like manner.
Its multimodal capabilities have made it possible for GPT-4 to pass exams like GRE (Graduate Record Examination), LSAT, SATs, AP (all languages), Intro Sommelier, Certified Sommelier, Advanced Sommelier, and USNCO Local Section Exam with flying colors.
Conclusion
As we have seen, GPT-4 is a language model that can do it all. With its advanced capabilities in processing visual inputs and generating human-like text, GPT-4 has set new benchmarks in professional and academic examinations.
Although it still needs to work on its coding abilities, GPT-4 has showcased its potential in assisting developers with their coding needs.
With further development and refinement, GPT-4 could become a powerful tool for various industries, revolutionizing the way we process and analyze data.
