chips
the text units (words or parts of words) that the model processes. Chatgpt model inputs and outputs are tokenized for efficient computation.
ability of the model to perform tasks for which it has not been specifically trained based on its general knowledge.
One-step learning is giving the model one usages of telemarketing list example, while n-step learning is giving the model many examples to learn from.
Attention mechanism
component of the transformer model that allows you to focus on different parts of the input text when generating responses.
Hallucination
An AI model "hallucinates" when it generates incorrect or meaningless information. Hallucinations can be mitigated with strategies such as augmented recovery generation (rag).
Chain reasoning
a method that helps the model think step by step, improving its ability to handle complex prompts or tasks.
Some chatgpt models are automatically equipped with this strategy, such as the latest openai o1 models. But you can ask any version to do chain reasoning: just ask it to explain its reasoning step by step.
Previous training
initial phase in which the model is trained on a massive data set to learn linguistic patterns before refining it for specific tasks.
Tuning
process of refining the model on a more limited data set or task to improve its performance for specific use cases.
Context window
the limit on the amount of input text the model can consider when generating a response.
A low context window means you can't send a long report and ask for a summary: the model will have "forgotten" the beginning of the document.
How to customize chatgpt
There are several ways to customize the powerful llms, such as the gpt engine that powers chatgpt:
to measure gpts
Openai allows its users to customize gpts to their liking. You can instruct a custom gpt to help you learn the rules of a certain board game, design rock metal band posters, or teach you AI concepts.
Custom AI Agents
With the advancement of AI technology, it's easy (and free) to create your own LLM-powered AI agents.
From low-code drag-and-drop builders, to advanced coding ecosystems, there are great AI building platforms for any use case and skill level.
Creating your own llm-based agent means you can design a bespoke AI assistant that schedules your meetings and generates your weekly metrics reports. Or you can create a customer service AI agent that you deploy to WhatsApp. The possibilities are not lacking.
