Ok but calculators are only allowed in math class and if there’s one thing language models suck at, it’s doing basic math. Forget anything at least as complicated as algebra
Dude, I was wondering why someone hadn’t done this the moment they discovered Ai was terrible at math. I would have imagine the crowds who deal with both would have some overlap at least.
For me they weren’t allowed in Calc I, II, III, Alg I, II and Differential equations. Every other class pretty much required it.
if there’s one thing language models suck at, it’s doing basic math.
If you’re using a GPT 3.5 turbo level models, sure. Synthetic data is perfect for teaching LLMs, o1 will be good enough up to Calc III IMO, maybe even better.
The only thing I don’t like about this is that it uses a TI, yikes.
LLMs do suck at math, if you look into it, the o1 models actually escape the LLM output and write a python function to calculate the output, I’ve been able to break their math functions by asking for functions that use math not in the standard Python library.
I know someone also wrote a wolfram integration to help solve LLMs math problems.
Terrence Tao (one of the most famous and active mathematician) recently wrote his thoughts in Mastodon on o1 mathematical capabilities. Interesting read: https://mathstodon.xyz/@tao/113132502735585408
Thanks for sharing, knew him from some numberphile vids cool to see they have a mastadon account. Good to know that LLMs are crawling from “incompentent graduate” to “mediocre graduate”. Which basically means its already smarter than most people for many kinds of reasoning task.
I’m not a big fan of the way the guy speaks though, as is common for super intelligent academic types they have to use overly complicated wording to formally describe even the most basic opinions while mixing in hints of inflated ego and intellectual superiority. He should start experimenting with having o-1 as his editor and summarize his toots.
Yeah, well, TI has spent bucketloads of money bribing textbook publishers to only include instruction for their specific models so they are now the de facto standard in American schools. This is apparently legal.
That’s also what my upper level math courses were like in college. In high school and the first couple of years of college I got good use out of my graphing calculator, but after that I reached the point where all of its advanced features were no longer useful. I just ended up leaving it at home and brought my old TI-30 Solar for class for the occasional time I had to crunch some actual numbers.
Don’t know about university math, but this applied to a lot of the stuff in my last years of school. Since we always had a part where you were required to solve everything without a calculator you had to be able to do everything without it. For algebra and Calculus it just meant that you were able to do the math more efficiently. For statistics the calculator was basically useless, since it didnt help you if you didnt knew what you had to do, what was basically the only hard part of it.
Ok but calculators are only allowed in math class and if there’s one thing language models suck at, it’s doing basic math. Forget anything at least as complicated as algebra
Time to build a wolfarm applet
Dude, I was wondering why someone hadn’t done this the moment they discovered Ai was terrible at math. I would have imagine the crowds who deal with both would have some overlap at least.
This kid never took science
You have an option not to take science?
Damn. Where I live the three main sciences were mandatory and they were all separate subjects
For me they weren’t allowed in Calc I, II, III, Alg I, II and Differential equations. Every other class pretty much required it.
If you’re using a GPT 3.5 turbo level models, sure. Synthetic data is perfect for teaching LLMs, o1 will be good enough up to Calc III IMO, maybe even better.
The only thing I don’t like about this is that it uses a TI, yikes.
LLMs do suck at math, if you look into it, the o1 models actually escape the LLM output and write a python function to calculate the output, I’ve been able to break their math functions by asking for functions that use math not in the standard Python library.
I know someone also wrote a wolfram integration to help solve LLMs math problems.
Wow that’s really clever actually. Basically using the library as digital scratch paper
Terrence Tao (one of the most famous and active mathematician) recently wrote his thoughts in Mastodon on o1 mathematical capabilities. Interesting read: https://mathstodon.xyz/@tao/113132502735585408
Thanks for sharing, knew him from some numberphile vids cool to see they have a mastadon account. Good to know that LLMs are crawling from “incompentent graduate” to “mediocre graduate”. Which basically means its already smarter than most people for many kinds of reasoning task.
I’m not a big fan of the way the guy speaks though, as is common for super intelligent academic types they have to use overly complicated wording to formally describe even the most basic opinions while mixing in hints of inflated ego and intellectual superiority. He should start experimenting with having o-1 as his editor and summarize his toots.
The language wasn’t that complex
Yeah, well, TI has spent bucketloads of money bribing textbook publishers to only include instruction for their specific models so they are now the de facto standard in American schools. This is apparently legal.
Anyway, team Casio represent.
Sorry, I’m an HP guy. I love their calculators, hate everything else they do, except their plotters maybe.
They let us use them for all my college math classes.
They really don’t help much at all if you don’t understand the math, and if you do understand, you don’t need the calculator most of the time.
That’s also what my upper level math courses were like in college. In high school and the first couple of years of college I got good use out of my graphing calculator, but after that I reached the point where all of its advanced features were no longer useful. I just ended up leaving it at home and brought my old TI-30 Solar for class for the occasional time I had to crunch some actual numbers.
Don’t know about university math, but this applied to a lot of the stuff in my last years of school. Since we always had a part where you were required to solve everything without a calculator you had to be able to do everything without it. For algebra and Calculus it just meant that you were able to do the math more efficiently. For statistics the calculator was basically useless, since it didnt help you if you didnt knew what you had to do, what was basically the only hard part of it.
They’re great at multiple choice when they’ve seen the test versions