Easy Embeddings Question 69 of 223

Do you embed the user query the same way you embed documents?

GenAI / LLM · Speak this in 60–90 seconds · Faridabad & Delhi NCR

Without indexScan every row

With indexJump to keys

CostWrites slower

Often yes with a symmetric embedder, but some models are asymmetric: one tower for queries and one for passages.

LLMs use tokens (not full words) because it helps them:

Embeddings questions separate people

who only read docs from people who shipped.

Keep it short, concrete,

and tied to GenAI / LLM work.

Stable token IDs

Each piece maps to a number the network can learn.

Fits the model

Fixed pieces are what transformers expect as input.

Before the model can read a sentence, it goes through these steps:

1
Often yes with a
symmetric embedder, but some models are asymmetric: one tower for queries and one for passages.
2
Using the wrong tower
hurts recall.
3
Read the embedding model's
documentation before indexing.
4
Context mix
Attention looks at nearby tokens together.
5
Next token
The model scores what should come next.
6
Decode
IDs turn back into readable text.

Let's see how a real sentence is tokenized (tokens may vary by model):

Input text

“Using the wrong tower hurts recall.”

Tokenized output

Usingthewrongtowerhurtsrecall

Token IDs (example)

2987408337471632900

Note: Actual tokens and IDs depend on the tokenizer (e.g., GPT, Llama, etc.).

Often yes with a symmetric embedder, but some models are asymmetric: one tower for queries and one for passages. Using the wrong tower hurts recall.