I've seen a lot of news lately about the ways that particular LLMs score on particular tests.

Which if any of those tests can I go take online to see how my performance on them compares to the models?

New Answer
Ask Related Question
New Comment

2 Answers sorted by

On twitter the IQ score of IIRC 84 for ChatGPT and 96 for GPT-4 were making the rounds, maybe you refer to those? I believe these scores are based on this freely available online test: 


I took it on wednesday just for fun. It's purely text-based but involves many different types of reasoning (including spatial reasoning). It's also a timed test which arguably inflates the LLM scores compared to humans. 

Is this one of those tests where you spend lot of time answering the questions, and at the end there is "if you want to see the results, send money"?

Also, is there any reason to believe that the test was actually somehow validated, as opposed to just numbers completely made up?

This is one of the tests where you spent not too much time (I think I took 13 minutes) and at the end there was "this is your result, if you want to see fine-grained scores, send money". Well, they claim that it is somehow validated and my score was somewhat realistic.  Main negative point was the need to provide an email for the result. 

Not for the all of them, but for the many of them you can see data and other info around here : https://paperswithcode.com/dataset/mmlu 

I browsed around but cannot find the actual mmlu questions, or an example of 1 question.  How do I view them>

"Homepage" button links to github, github readme links to tar with tests. Yeah, it's kinda not evident in some cases.
2 comments, sorted by Click to highlight new comments since: Today at 7:32 PM

Not sure if you've seen this or not: https://mashable.com/article/openai-gpt-4-exam-scores

But that references a number of standardized tests, some of which I suspect you have also taken. Here are a could of links to practice test that might have good matches for you to try.



[Rewrite as I don't think the first comment was actually helpful.]

Thank you! I didn't see your first version of this, but your current version is helpful for the human-specific tests that they're benchmarked on :)

New to LessWrong?