How I Used Gemini 3 to Turn a Research Paper into an Interactive Website in Just Three Minutes, Tripling My Reading Efficiency

Yesterday, as usual, I was reading news and academic research related to industry security, but some of the papers were really technical and I couldn’t keep going after a while.

I wondered if I could use a vibe coding tool to turn a research paper into an interactive, visual website, maybe even with built-in RAG and AI discussion features.

So I gave it a try. I picked a paper titled “Prompt Engineering vs. Fine-Tuning for LLM-Based Vulnerability Detection in Solana and Algorand Smart Contracts,” which mainly discusses whether prompt tuning or fine-tuning is better for helping large models find vulnerabilities. Using a one-shot prompt, I quickly created a really interesting website.

As you can see in the picture, the paper is nicely divided into several sections—there’s an overview, key findings (Deepseek performs better than Llama), and detailed arguments (such as what training data was used).

The website also includes built-in benchmark charts, showing the specific scores of the two models under different ecosystems and strategies.

On the right side, there’s an AI assistant ready to answer any questions. For example, when I asked what the F1 Score in the chart meant, the AI assistant quickly pulled information from the paper and gave me an explanation.

My takeaway is that vibe coding doesn’t have to be about building huge products. On the contrary, it’s perfect for quickly meeting our fleeting, specific needs. Things that used to require a lot of manual effort can now be easily handled by AI.

Vibe Coder isn’t a profession—it’s a lifestyle and a way of expressing yourself. The underlying idea is whether you want to live a more creative life.

Finally, here’s the website link:
Feel free to check it out and interact!