Over just a few months, ChatGPT went from correctly answering a simple math problem 98% of the time to just 2%, study finds. Researchers found wild fluctuations—called drift—in the technology’s abi...

L4sBot · 1 year ago

Over just a few months, ChatGPT went from correctly answering a simple math problem 98% of the time to just 2%, study finds. Researchers found wild fluctuations—called drift—in the technology’s abi...

@teydam · 1 year ago

This is peer-reviewed? they use a line in the discussion which seems relatively unprofessional, telling people to join a 12-step program if they like to use artificial training data.

@[email protected] · 1 year ago

I think arvix has no rule requiring a paper be per reviewed before uploading.

@[email protected] · edit-2 23 days ago

deleted by creator

@[email protected] · 1 year ago

Not affiliated with the paper in any way. Have just been following the news around it.

@blue_zephyr · edit-2 1 year ago

ArXiv papers are never peer reviewed.

@teydam · 1 year ago

Thank you

Over just a few months, ChatGPT went from correctly answering a simple math problem 98% of the time to just 2%, study finds. Researchers found wild fluctuations—called drift—in the technology’s abi...

Over just a few months, ChatGPT went from correctly answering a simple math problem 98% of the time to just 2%, study finds. Researchers found wild fluctuations—called drift—in the technology’s abi...

ChatGPT can get worse over time, Stanford study finds | Fortune