Re: [sympy] Re: ChatGPT and SymPy

It is time to integrate ChatGPT to SymPy! Elimboto On Mon, Jul 17, 2023 at 6:19 PM Aaron Meurer wrote: > I'm hopeful that this will. It seems to automatically go for SymPy > whenever faced with a math problem, or at least one that uses symbolic math > (it also uses NumPy when it can). The only

Re: [sympy] Re: ChatGPT and SymPy

I'm hopeful that this will. It seems to automatically go for SymPy whenever faced with a math problem, or at least one that uses symbolic math (it also uses NumPy when it can). The only downside is that the code interpreter hides the code by default, so people won't be able to see that it is using

Re: [sympy] Re: ChatGPT and SymPy

Nice. Are there any chances that ChatGPT may add visibility to SymPy? I feel like a lot of people who could be potential users of SymPy don't know of SymPy's existence. On Monday, July 10, 2023 at 4:35:00 p.m. UTC+2 Aaron Meurer wrote: > Interesting update on this. I asked ChatGPT with code

Re: [sympy] Re: ChatGPT and SymPy

The only way you always guarantee get right answer is to provide the ChatGPT two correct knowledge, but ask it to pick the better one. For solving the math problems, I think that these sort of encoding may be helpful. On Friday, March 31, 2023 at 7:23:41 AM UTC+9 asme...@gmail.com wrote: >

Re: [sympy] Re: ChatGPT and SymPy

It's the same as the example I showed above. It basically can't do math (or really, it can't even do arithmetic). It seems to have an especially hard time with fractions. I asked it to generate code using Python to compute the norm or [1/2, 1, 1]. It was able to generate accurate code using pure

Re: [sympy] Re: ChatGPT and SymPy

Use with caution -- the set-up can be right and the answer wrong as in: What is the norm of vector (1/2, 1, 1)? The norm of a vector is defined as the length or magnitude of the vector. The norm of a vector v = (v1, v2, ..., vn) is given by the formula: ||v|| = sqrt(v1^2 + v2^2 + ... + vn^2)

Re: [sympy] Re: ChatGPT and SymPy

ChatGPT is quite impressive! For those who do not have access to ChatGPT and/or Python, can use it with SageCell at https://tssfl.com/viewtopic.php?t=6651 SageCell supports countless Python libraries including SymPy. I tested ChatGPT with a number of examples, it tries solve even complex maths

Re: [sympy] Re: ChatGPT and SymPy

If you don't have access to the ChatGPT beta, someone made a Streamlit app that does the same thing. You can give it a math problem and it uses ChatGPT to generate and run SymPy code to solve it https://mathgpt.streamlit.app/ Aaron Meurer On Thu, Mar 23, 2023 at 3:28 PM Aaron Meurer wrote: >

Re: [sympy] Re: ChatGPT and SymPy

Actually I didn't notice it but there's literally an example of ChatGPT generating SymPy code and executing on https://openai.com/blog/chatgpt-plugins (scroll down to where it says "code interpreter"). It's their main example of calling out to Python. Aaron Meurer On Thu, Mar 23, 2023 at 2:42 PM

Re: [sympy] Re: ChatGPT and SymPy

On Thu, Mar 23, 2023 at 12:24 PM S.Y. Lee wrote: > Wolfram had recently announced the collaboration of chatGPT and wolfram > alpha > ChatGPT Gets Its “Wolfram Superpowers”!—Stephen Wolfram Writings > I wouldn't

Re: [sympy] Re: ChatGPT and SymPy

Wolfram had recently announced the collaboration of chatGPT and wolfram alpha ChatGPT Gets Its “Wolfram Superpowers”!—Stephen Wolfram Writings They start to use chatgpt to generate the Wolfram code. And it is

Re: [sympy] Re: ChatGPT and SymPy

In reviewing a PR related to units, I found ChatGPT to get correct the idea that a foot is bigger than an inch, but it said that a volt is bigger than a statvolt (see quoted GPT response [here](https://github.com/sympy/sympy/pull/24325#issuecomment-1354343306)). /c On Thursday, December 15,

Re: [sympy] Re: ChatGPT and SymPy

The trend with LLMs is much less structured. It doesn't use any formalism. It just guesses the next character of the input based on training on billions of examples. That's why I think that tools like SymPy that are more structured can be useful. GPT can already write SymPy code pretty well, much

Re: [sympy] Re: ChatGPT and SymPy

> My hope is that tools like SymPy can be used as oracles for tools like GPT to help them verify their mathematics. In the most general context, "correct mathematics" can also be considered some "grammar". So there should be some grammar between Type-0 grammar to Type-1 grammar in Chomsky

[sympy] Re: ChatGPT and SymPy

Just for some additional perspective. I have also tried this on some general chemistry word problems. In general, I see it getting basic one logical step processes correct (e.g. a single step dilution or grams -> moles). Things with multiple steps or requiring understanding the physical

[sympy] Re: ChatGPT and SymPy

[image: chatgpt.sympy.matrix_diag.png] On Wednesday, December 14, 2022 at 11:26:37 p.m. UTC+1 Francesco Bonazzi wrote: > Not everything is perfect... ChatGPT misses the *convert_to( ... ) *function > in *sympy.physics.units*, furthermore, the given code does not work: > > [image:

[sympy] Re: ChatGPT and SymPy

Not everything is perfect... ChatGPT misses the *convert_to( ... ) *function in *sympy.physics.units*, furthermore, the given code does not work: [image: chatgpt.sympy.unit_conv.png] On Wednesday, December 14, 2022 at 11:24:29 p.m. UTC+1 Francesco Bonazzi wrote: > [image:

[sympy] Re: ChatGPT and SymPy

[image: chatgpt.sympy.logical_inference.png] On Wednesday, December 14, 2022 at 11:23:43 p.m. UTC+1 Francesco Bonazzi wrote: > https://en.wikipedia.org/wiki/ChatGPT > > Some tested examples attached as pictures to this post. Quite impressive... > > -- You received this message because you are