Positional Encoding

Vision Transformer (ViT)

भिजन ट्रान्सफर्मर

A transformer architecture applied to images by treating image patches as tokens.

Activation Function

एक्टिभेसन फंक्सन

A function that decides whether and how strongly a neuron fires, adding non-linearity.

Autoregressive

अटोरिग्रेसिभ

Generating output one token at a time, each based on all the tokens produced before it.

Autoscaling

अटोस्केलिङ

Automatically adding or removing computing resources as demand changes.

Bag of Words

ब्याग अफ वर्ड्स

A simple text representation counting word occurrences while ignoring order.

BERT

बर्ट (BERT)

An influential transformer model that reads text in both directions to understand context.

Byte-Pair Encoding (BPE)

बाइट-पेयर इन्कोडिङ

A tokenization method that builds subword units from frequent character pairs.

Channel

च्यानल

A separate layer of information in data, such as the red, green, or blue in an image.