Keep both README images and link the website GitHub logo to the repository

README.md

@@ -8,7 +8,7 @@
 
 [Python notebook](https://colab.research.google.com/github/Affirm/tabular-foundation-models-tutorial/blob/main/materials/notebooks/01_tabicl_primer.ipynb) · [Our guide](#our-interactive-guide) · [Model landscape](#model-landscape) · [Learning resources](#tutorials-and-learning-resources) · [Conference resources](#conference-and-workshop-resources)
 
-Table as Prompt: An Interactive Guide to Tabular Foundation Models - accepted for presentation at the [NeurIPS 2026 Education Track](https://neurips.cc/Conferences/2026/CallforEducationalResources). [Read the paper (PDF, submission version)](materials/website/paper.pdf).
+**Table as Prompt: An Interactive Guide to Tabular Foundation Models** - accepted for presentation at the [NeurIPS 2026 Education Track](https://neurips.cc/Conferences/2026/CallforEducationalResources). [Read the paper (PDF, submission version)](materials/website/paper.pdf).
 
 ## What are tabular foundation models?
 
@@ -20,6 +20,8 @@ The broader goal is reusable prediction for structured data. The field now cover
 
 ## Our interactive guide
 
+![Tabular Foundation Models Tutorial: Interactive Guide, Papers and Code, and Benchmarks. Tabular in-context learning uses labeled examples and a new row's features as inputs to a pretrained Transformer with fixed weights to predict the new row's label.](assets/tabular-tutorial-overview.png)
+
 Open the [guide](https://affirm.github.io/tabular-foundation-models-tutorial/) in a modern browser. Follow the sequence from tabular prediction and adaptation through PFN theory, TabICLv2 training, and inference. Then try the browser playground and inspect the model's intermediate computations. Basic supervised learning is enough to get started.
 
 TabICLv2 is the worked example. The guide distinguishes explanatory simulations from real model execution; browser inference runs locally. For a Python exercise, use the [notebook](materials/notebooks/01_tabicl_primer.ipynb) or [open it in Colab](https://colab.research.google.com/github/Affirm/tabular-foundation-models-tutorial/blob/main/materials/notebooks/01_tabicl_primer.ipynb). [Setup instructions](materials/README.md) cover the notebook environment.

materials/website/index.html

@@ -34,11 +34,11 @@
         <span class="mark__dot"></span>
         <span class="mark__name">NeurIPS 2026 · Education Track</span>
       </a>
-      <a class="github-link mono" href="https://affirm.github.io/tabular-foundation-models-tutorial/" target="_blank" rel="noopener" aria-label="Open the public tutorial website">
+      <a class="github-link mono" href="https://github.com/Affirm/tabular-foundation-models-tutorial" target="_blank" rel="noopener" aria-label="Open the tutorial GitHub repository">
         <svg viewBox="0 0 24 24" aria-hidden="true">
           <path fill="currentColor" d="M12 .7a11.5 11.5 0 0 0-3.64 22.4c.58.1.79-.25.79-.56v-2.23c-3.22.7-3.9-1.37-3.9-1.37-.53-1.34-1.29-1.7-1.29-1.7-1.05-.72.08-.71.08-.71 1.17.08 1.78 1.2 1.78 1.2 1.04 1.78 2.72 1.27 3.39.97.1-.75.4-1.27.74-1.56-2.57-.29-5.27-1.28-5.27-5.69 0-1.26.45-2.28 1.19-3.09-.12-.29-.52-1.47.11-3.05 0 0 .97-.31 3.16 1.18a10.9 10.9 0 0 1 5.76 0c2.2-1.49 3.16-1.18 3.16-1.18.63 1.58.23 2.76.11 3.05.74.81 1.19 1.83 1.19 3.09 0 4.42-2.71 5.39-5.29 5.68.42.36.79 1.06.79 2.14v3.18c0 .31.21.67.8.56A11.5 11.5 0 0 0 12 .7Z"/>
         </svg>
-        <span>Website</span>
+        <span>GitHub</span>
       </a>
     </div>
     <a class="bar__page-title" href="#top">Tabular Foundation Models</a>