LLM's not MLM's
The company that introduced the first GPT model in 2018.
What is OpenAI?
A type of neural network that processes input sequences to create output sequences
What is a transformer?
What is Temperature?
Teaches algorithms to make decisions by trial and error, while generative AI creates new data points based on patterns in existing data but can work together with generative AI to creative more effective AI systems
A type of AI that creates new content, such as text, images, videos, and music. It uses generative models to analyze existing data and generate new data with similar characteristics.
What is Generative AI?
These enable the engine behind modern AI applications and power everything from code generation to content creation
What are LLM's?
What is Prompt Engineering?
The process of identifying and reducing biases in artificial intelligence
What is bias mitigation?
A subfield of artificial intelligence that allows computers to understand, interpret, and generate human language
What is Natural Language Processing?
A mathematical formula that measures the difference between the outputs generated by the model (like text, images, or audio) and the desired "ground truth" data.
What is loss function?
Customization for specific use cases to create specialized AI solutions
What is fine-tuning?
The smallest units of data that AI models process. They can be characters, words, phrases, or subwords.
What are tokens?
A parameter whose value is set before the machine learning process begins. In contrast, the values of other parameters are derived via training
What is hyper parameter?
A statistical method used to assess the performance of a machine learning model by repeatedly dividing a dataset into training and testing subsets
What is Cross-validation?
The process of using a trained model to create new output data based on input data.
Examples include: Computer vision, Spam filtering, Speech recognition, Document summarization
What is inference?
Numerical representations of data that are used in generative AI (GenAI) to capture relationships between inputs. Powers semantic search and recommendations by making AI understand similarity and context.
What is embeddings?
Improves accuracy and explainability by enabling complex problem-solving capabilities with step-by-step reasoning processes
What is chain-of-thought?
Reduces training requirements by enabling flexibility in AI applications
What is zero shot learning?
A technique where multiple generative models are combined to produce a more robust and accurate output, leveraging the strengths of each individual model to overcome their weaknesses, resulting in a better overall prediction than any single model alone; essentially, it's like consulting multiple AI "experts" to get a more comprehensive answer.
What is ensemble learning?
When a model is trained too well on a specific set of data, making it difficult to predict new data
What is overfitting?
What is RAG or Retrieval Augmented Generation?
Determines task complexity handling by processing task capacity with document chunking strategies, memory management and conversation design
What is context window?
A technique that helps machines focus on the most important parts of an input sequence
What is attention mechanism?
A process that uses AI to improve machine learning models by selecting, transforming, and generating features from raw data
What is feature engineering?
When a generative AI model produces inaccurate, false, or misleading results
What is a hallucination?