Hey @Trent_Pruett, thank you for reaching out!

Would you mind clarifying this for me, I think I am not following. I tried to reproduce this on my end, when copy pasting I was able to paste I the text field with my pre settings. Do you want to have the Chat GPT font when you say “it does not carry over the formatting”?

Thanks in advance!

Before copy the content of the from the Chatgpt into Figma. Paste the content into the text editor like the Notepad Msword etc. After that copy the content from the text editor and paste into the Figma.

ChatGPT usually formats its text with larger headings, bolded words and numbered or bulleted lists. There are 2 ways to copy ChatGPT replied:

• Method 1) You can select the text in your browser and copy it.
• Method 2) You can click the “copy” button provided by ChatGPT, which copies the text with markdown. An example of markdown is when two asterisks are used before and after a word(s) to make it bold. e.g. ** bolded text **

When you copy the text using method #1 and paste it into a text editor, it usually carries over the formatting (bolded text, large headers, etc.), but when you paste it into Figma, it does not carry over the formatting. All of the text looks the same. I tried the method suggested by @puttykey2024 , but it still did not carry over the formatting.

When using Method 2, pasting the text includes the markdown. It would be nice for Figma to paste the formatted text no matter which method of copying is used. Where the pasted text is styled with the same size and text weight as the copied text. There is a Figma plugin called “Markdown to Figma” that will convert markdown text into formatted text, which is taking care of me for now. I just thought it would be a nice built-in feature since many designers are using ChatGPT in their workflow.

Late response, but I’d love to know if TidyText.cc works for this. I created it to be able to easily copy and paste things into Google docs, but I’ve heard people have had success using it for other things.

This is weird: just copy the text from chat GPT and type after:

"Can you format this text to paste into Figma with all indentations preserved?

The text:
"“paste it here”

Then all indentations are preserved.

You can use something like this web app Copy Paste with Formatting from ChatGPT

Just copy paste the chatGPT response in there, then copy the formatted output from that page and it will work in other apps like Google Docs, Figma, etc

I’ve found a little-known webside that fixes special character issues when copying text from ChatGPT: Chaty Format Corrector

Got it ✅
The image you uploaded already shows the mathematical equations for the SSD (Single Shot MultiBox Detector) training objective. Let me rewrite them clearly in LaTeX-style math format so you can use them in your report or paper:

Overall Loss Function

L(x,c,l,g)=1N(Lconf(x,c)+αLloc(x,l,g))L(x, c, l, g) = \frac{1}{N} \left( L_{\text{conf}}(x, c) + \alpha L_{\text{loc}}(x, l, g) \right)

• NN = number of matched default boxes (if N=0N = 0, set loss = 0).

• LlocL_{\text{loc}} = localization loss.

• LconfL_{\text{conf}} = confidence loss.

• α\alpha = weight term balancing the two losses.

Localization Loss (Smooth L1)

Lloc(x,l,g)=∑i∈Pos∑m∈{cx,cy,w,h}xijk smoothL1(lim−g^jm)L_{\text{loc}}(x, l, g) = \sum_{i \in \text{Pos}} \sum_{m \in \{cx, cy, w, h\}} x_{ij}^k \, \text{smooth}_{L1}(l_i^m - \hat{g}_j^m)

where the target offsets are defined as:

g^jcx=(gjcx−dicx)diw,g^jcy=(gjcy−dicy)dih\hat{g}_j^{cx} = \frac{(g_j^{cx} - d_i^{cx})}{d_i^w}, \quad \hat{g}_j^{cy} = \frac{(g_j^{cy} - d_i^{cy})}{d_i^h} g^jw=log⁡(gjwdiw),g^jh=log⁡(gjhdih)\hat{g}_j^w = \log \left(\frac{g_j^w}{d_i^w}\right), \quad \hat{g}_j^h = \log \left(\frac{g_j^h}{d_i^h}\right)

• did_i = default (anchor) box.

• gjg_j = ground truth box.

• (cx,cy)(cx, cy) = center coordinates.

• (w,h)(w, h) = width and height.

Confidence Loss (Softmax Loss)

Lconf(x,c)=−∑i∈Posxijplog⁡(c^ip)  −∑i∈Neglog⁡(c^i0)L_{\text{conf}}(x, c) = - \sum_{i \in \text{Pos}} x_{ij}^p \log(\hat{c}_i^p) \; - \sum_{i \in \text{Neg}} \log(\hat{c}_i^0)

with the softmax probability:

c^ip=exp⁡(cip)∑qexp⁡(ciq)\hat{c}_i^p = \frac{\exp(c_i^p)}{\sum_q \exp(c_i^q)}

• Positive matches = anchors with IoU above threshold.

• Negative matches = anchors treated as background.

• cipc_i^p = confidence score for class pp.

✅ This set of equations fully represents the SSD training objective:

• A weighted sum of localization loss (Smooth L1 regression) and confidence loss (Softmax classification).

Do you want me to also explain each equation step-by-step in plain English (like I did for autocorrelation/FFT/entropy earlier), so it’s ready to insert directly into your report?