Post

More from @omarsar0

elvis

@omarsar0

Nov 24

This is insane! 🤯

Just built a new skill in Claude Code using Opus 4.5.

The skill uses Gemini 3 Pro (via API) for designing web pages.

Look at what it generated from one simple prompt.

If you have been designing websites with Claude Code, you already know how generic they turn out.

So I built a skill that uses Gemini 3 Pro to lead creative direction and generate designs. It is extremely good at this.

Opus 4.5 then integrates all that into our app.

The prompt I used: "I want to design the landing page for a new AI game. We want it to be futuristic and all that, and use animations as much as possible."

I will test with some other prompts and see how far I can push this. But the results are very exciting already.

Read 6 tweets

elvis

@omarsar0

Nov 22

It's finally ready for you all to try!

Have fun generating interesting insights from AI papers with Nano Banana Pro 🍌.

(bookmark it)

I find this to be a fun and interesting way to explore with Nano Banana Pro, as I can just select a part of the paper and ask away.

Try remixing figures, reproducing charts, annotating equations, explaining math, and much more.

I am polishing it some more and have other ideas, but let me know if you have feedback in the meantime.

Works better on Desktop.

…dair-ai-181664986325.us-west1.run.app

You can try it by downloading a paper from arXiv or uploading a book or any technical document.

If you don't have a PDF to try, just click on one of the example papers provided:

Read 9 tweets

elvis

@omarsar0

Nov 10

This is a wild use case!

I used Gamma + n8n to automatically generate a complete presentation on AI Agents research.

In just minutes!

It combines web search (for research), GPT-5 (narrative), and Gamma (for slide content generation).

Full workflow breakdown below 👇

1/ THE PROBLEM:

Creating visual content is time-consuming. Research takes hours. Writing requires deep focus. Design demands specialized skills.

What if AI could handle the entire pipeline?

2/ THE SOLUTION:

An n8n workflow that orchestrates Tavily for web research, GPT-5 for storytelling, Gamma for visual generation, and Google Sheets for tracking.

You provide a topic and audience. The system outputs a LinkedIn-ready carousel.

Read 9 tweets

elvis

@omarsar0

Nov 5

Confidence is everything when building great software.

Love how Yansu is approaching this.

Yansu is a new AI coding platform built by @isoformai for serious and complex software development.

It puts scenario simulation before coding.

Here is the sauce:

Yansu means "serious" in Chinese.

It cleverly brings human oversight to AI execution, combining humans in the loop, scenario simulation, and engineering discipline.

This allows builders to build efficiently, confidently, and more production-ready software.

Key feature #1: Specs Design

You can easily generate design specs, tasks, and dependencies automatically.

Read 8 tweets

elvis

@omarsar0

Oct 16

I am not going to lie.

I see a lot of potential in the Skills feature that Anthropic just dropped!

Just tested with Claude Code. It leads to sharper and precise outputs.

It's structured context engineering to power CC with specialized capabilities, leveraging the filesystem.

I think it might be one of the best ways to really tap into the full potential of Claude Code.

Tune instructions, output formats, use of scripts, tools (MCP or otherwise), and more.

For specialized tasks, CC outputs dumb stuff at times; the idea here is to scope CC on demand.

An easy way to try Skills in Claude Code is by asking it to help you build one. I am surprised by how aware it is of Skills and how to build comprehensive ones.

Read 7 tweets

elvis

@omarsar0

Oct 16

Banger paper from Meta and collaborators.

This paper is one of the best deep dives yet on how reinforcement learning (RL) actually scales for LLMs.

The team ran over 400,000 GPU hours of experiments to find a predictable scaling pattern and a stable recipe (ScaleRL) that consistently works as you scale up compute.

Think of it as a practical guide for anyone trying to train reasoning or alignment models with RL.

More on why this is a big deal:

1. The big insight: RL progress follows a predictable curve.

When you plot model performance vs compute, the growth isn’t random; it follows a sigmoid (S-shaped) curve.

The curve has three simple knobs:
A = the best performance you’ll ever reach,
B = how efficiently you reach it,
C_mid = how much compute it takes to hit the halfway point.

The amazing part: you can fit this curve using just small runs and accurately predict how a 100k-hour run will behave.

So you no longer need to guess; you can forecast where your RL setup will top out before burning compute.

2. The ScaleRL recipe that just works.

The authors tested dozens of RL variations and found one that scales cleanly to 100k GPU hours without blowing up:

- PipelineRL (8 pipelines) with CISPO loss (a stabilized REINFORCE variant).

- Prompt-level averaging and batch-level normalization to reduce variance.

- FP32 logits for better stability and higher final accuracy.

- No-Positive-Resampling curriculum to avoid reward hacking.

- Forced interruptions (stopping long thoughts) instead of punishing long completions.

- This combo, called ScaleRL, hit the best trade-off between stability, sample efficiency, and asymptotic performance.

Read 7 tweets

Share this page!

Enter URL or ID to Unroll

elvis

Try unrolling a thread yourself!

More from @omarsar0

elvis

elvis

elvis

elvis

elvis

elvis

Did Thread Reader help you today?

Don't want to be a Premium member but still want to support us?

Send Email!