AI terms starting with T.

t-SNE

टी-एसएनई

A technique for visualizing high-dimensional data in two or three dimensions.

Tanh

टान्ह

An activation function outputting values between -1 and 1, centered at zero.

Target Variable

लक्ष्य चर

The value a model is trying to predict.

Technological Singularity

प्राविधिक एकत्व

A hypothetical future point when AI surpasses human intelligence and rapidly self-improves.

Temperature

टेम्परेचर

A setting that controls the randomness of a model's output; higher means more varied.

Tensor

टेन्सर

A multi-dimensional array of numbers that is the basic data structure in deep learning.

TensorFlow

टेन्सरफ्लो

An open-source machine learning framework developed by Google.

Test Set

परीक्षण सेट

Held-out data used to evaluate how well a trained model performs on unseen examples.

Text Classification

पाठ वर्गीकरण

Assigning a category to a piece of text, such as spam or not spam.

Text Summarization

पाठ संक्षेपण

Automatically producing a shorter version of a text that keeps the key points.

Text-to-Audio

पाठ-देखि-अडियो

Generating sound effects, speech, or music from a written description.

Text-to-Image

पाठ-देखि-तस्बिर

Generating images from a written description using AI.

Text-to-Speech (TTS)

पाठ-देखि-बोली

Converting written text into natural-sounding spoken audio.

Text-to-Video

पाठ-देखि-भिडियो

Generating short video clips from a written description using AI.

TF-IDF

टीएफ-आईडीएफ

A weighting that highlights words important to one document relative to a whole collection.

Threshold

थ्रेसहोल्ड / सीमा मान

A cutoff value that decides which class a prediction falls into.

Thresholding

थ्रेसहोल्डिङ

Converting an image to black and white by a brightness cutoff to separate objects.

Throughput

थ्रुपुट

The number of requests or items a system can process in a given time.

Time Series

समय शृङ्खला

Data points collected in time order, such as daily prices or hourly temperatures.

Token

टोकन

A small unit of text — a word or part of a word — that a language model processes.

Token Streaming

टोकन स्ट्रिमिङ

Sending a model's response token by token as it is generated, for faster feedback.

Tokenization

टोकनाइजेसन

The process of splitting text into tokens that a model can read.

Top-k Sampling

टप-के स्याम्पलिङ

Picking the next word from only the k most likely options for controlled variety.

Top-p (Nucleus) Sampling

टप-पी स्याम्पलिङ

Choosing the next word from the smallest set of options whose probabilities add up to p.

Topic Modeling

विषय मोडेलिङ

Discovering the abstract themes that run through a collection of documents.

Total Cost of Ownership

कुल स्वामित्व लागत

The full cost of an AI system over its life, including build, run, and maintenance.

Toxicity

विषाक्तता

Harmful, hateful, or abusive content that AI systems should detect and avoid producing.

TPU

टीपीयू

A specialized chip built by Google to accelerate machine learning workloads.

Training

तालिम / प्रशिक्षण

The process of feeding data to a model so it can learn the patterns needed to perform a task.

Training Set

तालिम सेट

The portion of data used to teach the model.

Transcription

प्रतिलेखन

Converting spoken audio into written text.

Transfer Learning

ट्रान्सफर लर्निङ

Reusing a model trained on one task as a starting point for a related task.

Transformer

ट्रान्सफर्मर

A neural network architecture using attention that powers most modern language models.

Transparency

पारदर्शिता

Making how an AI system works and makes decisions understandable to people.

Turing Test

ट्युरिङ परीक्षण

A test proposed by Alan Turing to judge whether a machine's responses are indistinguishable from a human's.