refactor: remove blog posts section from homepage and update research tasks and '5-dollar-llm' blog post content.

Author: vukrosic

RESEARCH.md

@@ -1,8 +1,11 @@
-### Open Tasks
+---
+
+## Open Tasks
 
 [Does RoPE mess with semantics of the vectors, what would you do differently? ➝](/blog/rope-semantics)
 
-Setup research environment for [Back to Basics: Let Denoising Generative Models Denoise](https://arxiv.org/pdf/2511.13720). Add research questions, ideas, topics, etc.
+
+[Make LLM pretraining accessible for everyone ➝](/blog/5-dollar-llm)
 
 
 ---

app/page.tsx

@@ -87,41 +87,6 @@ export default function Home() {
             )}
           </div>
 
-          {/* Dynamic Blog Posts Section */}
-          {posts.length > 0 && (
-            <div className="mt-24 border-t border-white/10 pt-16">
-              <h2 className="text-3xl font-bold text-[#faf9f6] mb-12 flex items-center gap-3">
-                <span className="text-2xl">📚</span>
-                Latest Research Articles
-              </h2>
-              <div className="space-y-12">
-                {posts.map((post) => (
-                  <div key={post.slug} className="group relative">
-                    <div className="flex items-center gap-4 text-sm text-[#faf9f6]/50 mb-3">
-                      <span>{post.date}</span>
-                    </div>
-                    <Link href={`/blog/${post.slug}`} className="block">
-                      <h3 className="text-2xl font-bold text-white mb-3">
-                        {post.title}
-                      </h3>
-                    </Link>
-                    <p className="text-[#faf9f6]/75 text-lg leading-relaxed mb-4">
-                      {post.description}
-                    </p>
-                    <Link
-                      href={`/blog/${post.slug}`}
-                      className="inline-flex items-center gap-2 text-blue-400 hover:text-blue-300 font-medium transition-colors"
-                    >
-                      Read full article
-                      <svg className="w-4 h-4 group-hover:translate-x-1 transition-transform" fill="none" stroke="currentColor" viewBox="0 0 24 24">
-                        <path strokeLinecap="round" strokeLinejoin="round" strokeWidth={2} d="M17 8l4 4m0 0l-4 4m4-4H3" />
-                      </svg>
-                    </Link>
-                  </div>
-                ))}
-              </div>
-            </div>
-          )}
         </div>
       </main>
     </>

blog-posts/5-dollar-llm.md

@@ -1,13 +1,11 @@
 ---
-title: "Pretrain GPT-5 on 1 GPU"
+title: "Pretrain GPT-2 on 1 GPU"
 date: "2026-01-19"
-description: "Research for maximum LLM training efficiency"
+description: "Make LLM pretraining accessible for everyone"
 ---
 
-Researching LLM to make it accessible to anyone.
+This [GitHub repo](https://github.com/Open-Superintelligence-Lab/5-dollar-llm) in an LLM that is accessible for anyone to pretrain. We aim to make LLM pretraining more and more accessible to everyone (not less and less, as common false expectation suggests).
 
-[github](https://github.com/Open-Superintelligence-Lab/5-dollar-llm)
+Currently GPT-1 has already be replicated in this repo. For GPT-2 we just need to increase the data size by 10x, but some architectural improvements would be good.
 
-
-
-![Mamba S6 Architecture Diagram](/content/mamba-deep-dive/architecture.png)
+Suggest your ideas, we need to empirically measure and validate improvements as described in the repo.