Scribd
Upload
39 views8 pages

Large Language Models: Foundation of

The document explains the foundational role of vectors and embeddings in AI, highlighting how vectors represent data in a format that algorithms can process. It discusses the concept of embeddings, which convert discrete data into numerical vectors, enabling mathematical operations for tasks like clustering and classification. Additionally, it covers cosine similarity as a method for measuring the semantic similarity between vectors, crucial for advancements in natural language processing.

Uploaded by

sumitraja15
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
39 views8 pages

Large Language Models: Foundation of

The document explains the foundational role of vectors and embeddings in AI, highlighting how vectors represent data in a format that algorithms can process. It discusses the concept of embeddings, which convert discrete data into numerical vectors, enabling mathematical operations for tasks like clustering and classification. Additionally, it covers cosine similarity as a method for measuring the semantic similarity between vectors, crucial for advancements in natural language processing.

Uploaded by

sumitraja15
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

Foundation of Large

Language Models
How is Math the backbone of AI?

(c) Copyrights Reserved [Link]


1

What are Vectors?

A vector is a mathematical point that represents data in a


format that AI algorithms can understand.

Vectors are arrays (or lists) of numbers, with each


number representing a specific feature or attribute of the
data.

(c) Copyrights Reserved [Link]


2

Vector
Representation
Let's say we have a 2D space where each axis
represents a taste characteristic: sweet and sour.

As we add more dimensions (e.g., bitter, salty, savory),


we get a richer understanding of the dishes.

The closer the vectors, the more similar their meanings.

(c) Copyrights Reserved [Link]


3

What are
Embeddings?
Embeddings convert words, phrases, or even images
into numerical vectors.

These vectors map discrete data into a continuous latent


space, capturing relationships.

(c) Copyrights Reserved [Link]


4

What do Embeddings
represent?
Embeddings allow mathematical operations on text,
making tasks like clustering, classification, and
regression possible.

By converting text to vectors, models can interpret,


compare, and manipulate words more effectively.

(c) Copyrights Reserved [Link]


5

Cosine similarity for


vector similarity
Cosine similarity is a measurement that quantifies the
similarity between two or more vectors.

It is the cosine of the angle between vectors, which are


typically non-zero and within an inner product space.

Source: Wikipedia

(c) Copyrights Reserved [Link]


6

Cosine similarity for


vector similarity
Cosine similarity is a measure in LLMs for evaluating the
semantic similarity between embeddings.

It is used in semantic search as it provides a robust and


efficient way to compare textual data based on its
meaning, driving significant advancements in NLP and
AI-driven applications.

(c) Copyrights Reserved [Link]


To learn more, join us on 15
January at 10AM PDT for
Simplifying Mathematics
Behind AI

Albar Wahab
Senior Data Scientist
Data Science Dojo

(c) Copyrights Reserved [Link]

You might also like