Skip to main content

Cannot copy LaTeX values

Feature RequestsChat

In the Venice UI it appears that it is using Mathjax for rendering math equations. Unfortunately there is no way for me to copy the LaTeX information because it is only rendered as an image like so:

<mjx-container class="MathJax" jax="SVG">…</svg></mjx-container>

This make it extremely hard to communicate with  the LLM. Imagine you are learning about economics and you see the nicely rendered equation “\dot{k} = s f(k) - (n + g + \delta) k”, but you don’t understand it so you want the LLM to define each term. 

There is no way of copying the latex information to easily paste back to the LLM!

I need, very much, the ability to copy the LaTeX code itself not just view a nicely rendered image of it. 

Please make the LaTeX copy-able any time it is rendered. Maybe by adding it to an attribute in the element? I want to be able to right click on any LaTeX and copy the LaTeX code itself. 

Thank you very much. 

3 comments

Log in to comment and vote

Comments3

  • Bronze Mist

    •

    Sep 29, 2025

    We can examine a similar example at 

    https://en.wikipedia.org/wiki/Black–Scholes_equation

    We can see that they both display the LaTeX as an image AND they have the LaTeX code in the alt attribute like so:

    <img src="https://wikimedia.org/api/rest_v1/media/math/render/svg/a316faebd9599d6d65e89d4f380956c9f0e96114" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -3.171ex; width:37.354ex; height:7.509ex;" alt="{\displaystyle {\begin{cases}{\frac {\partial V}{\partial t}}+{\frac {1}{2}}\sigma ^{2}S^{2}{\frac {\partial ^{2}V}{\partial S^{2}}}+rS{\frac {\partial V}{\partial S}}-rV=0\V(T,s)=K(s)\quad \forall s\end{cases}}}">

    Of course, in my case, I also want the UI functionality of being able to copy that LaTeX code value.

  • Bronze Mist

    •

    Sep 29, 2025

    This also raises a very important question: If you are not representing the LaTeX code anywhere, and it’s only an image, then what information is being submitted back to the LLM upon new chat discussion cycles?  Are you submitting the current conversation back to the LLM with missing LaTeX information, like Swiss cheese and the holes are the equations that the LLM previously output? If so that does not seem good to me. I want the LLM to have all the informational context of the conversation upon every submission. And further, if this is so, that I consider this to be a very serious issue. Please address as quickly as possible.

  • Bronze Mist

    •

    Sep 29, 2025

    I’m specifically talking about the web app in browser. I’m not sure if the same applies in the app. thank you!