Moderate Transformers Question 79 of 223

What does the feed-forward network do in a transformer block?

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

PICTURE THIS: A SENTENCE BECOMES TOKENS

The model does not read letters like humans. It reads these pieces, then predicts the next one.

Simple meaning

After attention mixes information across tokens, a position-wise MLP transforms each token independently.

WHY — Tokens instead of words?

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

Transformers 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.

STEPS — What happens with tokens?

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

1
After attention mixes information
across tokens, a position-wise MLP transforms each token independently.
2
It is where much
of the parameter count lives.
3
Embeddings
Attention is the communication
4
the FFN is a
large per-token memory.
5
Next token
The model scores what should come next.
6
Decode
IDs turn back into readable text.

EXAMPLE — See it in action

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

Input text

“It is where much of the parameter count lives.”

Tokenized output

Itiswheremuchofthe

Token IDs (example)

2987408337471632900

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

Key takeaway

After attention mixes information across tokens, a position-wise MLP transforms each token independently. It is where much of the parameter count lives.