# @awomanindatascience on Instagram

- **Type:** Video
- **Original URL:** https://www.instagram.com/p/DdX3aAeT6c1
- **Gondola URL:** https://gondola.cc/posts/70951320-awomanindatascience-instagram
- **Thumbnail:** https://img.gondola.cc/tr:w-,h-,fo-auto/postThumbnails/25695e801a.jpg
- **Posted:** 2026-09-17T03:13:39.000+00:00
- **Account Owner:** Chinar Arora (@awomanindatascience) — https://gondola.cc/awomanindatascience

## Caption

AI models aren’t just answering questions anymore.
They’re spending more computation to reason before they answer.

But what actually makes a model a reasoning model?

I’m starting a new series: Build a Reasoning Model From Scratch — following Sebastian Raschka’s Build a Reasoning Model From Scratch.

In Day 1, we break down:

🧠 What “reasoning” actually means for an LLM
⚙️ The pretraining → post-training foundation
🔬 3 ways to improve reasoning:
→ Inference-time scaling
→ Reinforcement learning
→ Distillation
🤖 Why reasoning matters for AI agents
💰 And the trade-off: more reasoning can mean more tokens, latency, and cost

The goal isn’t just to understand reasoning models.

We’re going to build one. From scratch.

Follow along for Day 2. 👀

#ReasoningModels #LLMFromScratch #AIEngineering #MachineLearning #DeepLearning LLM AIEngineering ChainOfThought DeepSeekR1 BuildInPublic GenerativeAI AIAgents

## Stats

- **Views:** 0
- **Likes:** 27
- **Shares:** 0
- **Comments:** 2

## Tags

aiengineering, reasoningmodels, deeplearning, machinelearning, llmfromscratch

---
Copyright (c) Gondola