---
title: "Mohitm (@mohitm) | Hashnode"
description: "Data science professional with around 15 years experience, working to on Pricing and Media analytics practice for CPG companies. Currently building a micro-l..."
url: "https://hashnode.com/@mohitm"
---

@mohitm

# Mohitm

@mohitm · Bangalore, India · Joined August 2026

### [Why Your LLM Can't Count the R's in 'Strawberry'](https://monthlydatascience.hashnode.dev/why-your-llm-can-t-count-the-r-s-in-strawberry?utm_source=hashnode&utm_medium=feed)

Ask ChatGPT how many R's are in strawberry and it often gets it wrong. The reason isn't a bug — it's tokenization. Here's what's actually happening, with an interactive demo.

6 min read · [monthlydatascience.hashnode.dev](https://monthlydatascience.hashnode.dev/why-your-llm-can-t-count-the-r-s-in-strawberry?utm_source=hashnode&utm_medium=feed)

[1](https://hashnode.com/posts/why-your-llm-can-t-count-the-r-s-in-strawberry/6ae15bd87700189050d03c0f) [Oct 28](https://hashnode.com/posts/why-your-llm-can-t-count-the-r-s-in-strawberry/6ae15bd87700189050d03c0f)

### [Embedding is just a vector](https://monthlydatascience.hashnode.dev/embedding-is-just-a-vector?utm_source=hashnode&utm_medium=feed)

Token embeddings are rows of a matrix, looked up by an integer. What is the vector, why similarity uses angles, & why embedding layer is context-free

6 min read · [monthlydatascience.hashnode.dev](https://monthlydatascience.hashnode.dev/embedding-is-just-a-vector?utm_source=hashnode&utm_medium=feed)

[3](https://hashnode.com/posts/embedding-is-just-a-vector/6ac31c2ce5fdec84e1b8951b) [Oct 5](https://hashnode.com/posts/embedding-is-just-a-vector/6ac31c2ce5fdec84e1b8951b)

### [What PCA is actually doing to your data](https://monthlydatascience.hashnode.dev/what-pca-is-actually-doing-to-your-data?utm_source=hashnode&utm_medium=feed)

It only sees variance, it never sees your target

7 min read · [monthlydatascience.hashnode.dev](https://monthlydatascience.hashnode.dev/what-pca-is-actually-doing-to-your-data?utm_source=hashnode&utm_medium=feed)

[Sep 1](https://hashnode.com/posts/what-pca-is-actually-doing-to-your-data/6a964b1846119e8824fa4b4a)

### [Correlation Is Pairwise. Multicollinearity Isn't](https://monthlydatascience.hashnode.dev/correlation-is-pairwise-multicollinearity-isn-t?utm_source=hashnode&utm_medium=feed)

Almost every regression project I've seen in the last decade starts the same way. Load the data, df.corr(), plot the heatmap, scan for red squares, drop one variable from every pair above 0.8. Then we

7 min read · [monthlydatascience.hashnode.dev](https://monthlydatascience.hashnode.dev/correlation-is-pairwise-multicollinearity-isn-t?utm_source=hashnode&utm_medium=feed)

[Aug 23](https://hashnode.com/posts/correlation-is-pairwise-multicollinearity-isn-t/6a8a6b9784450835822d7d0b)