# @awomanindatascience on Instagram

- **Type:** Video
- **Original URL:** https://www.instagram.com/p/DIXy_AURMAe
- **Gondola URL:** https://gondola.cc/posts/61230118-awomanindatascience-instagram
- **Thumbnail:** https://img.gondola.cc/tr:w-,h-,fo-auto/postThumbnails/b991bc137e.jpg
- **Posted:** 2025-04-13T03:34:50.000+00:00
- **Account Owner:** Chinar Arora (@awomanindatascience) — https://gondola.cc/awomanindatascience

## Caption

It’s Day 8 of Building LLMs from Scratch:
Implemented self-attention, the core mechanism behind transformers and GPT models! This involves computing attention scores using scaled dot-product attention:
	1.	Project inputs into Query (Q), Key (K), and Value (V) matrices
	2.	Compute attention scores using Q · Kᵀ / sqrt(d_k)
	3.	Apply softmax to get attention weights
	4.	Multiply with V to get the weighted output

This lets the model focus on relevant parts of the input—like how each word relates to others in a sentence!#deeplearning #generativeai #30dayschallenge #womenwhocode #datascience #largelanguagemodels #womenintech

## Stats

- **Views:** 2,105
- **Likes:** 76
- **Shares:** 0
- **Comments:** 4

## Tags

womenwhocode, generativeai, datascience, womenintech, largelanguagemodels, deeplearning, 30dayschallenge

---
Copyright (c) Gondola