conversational assistant

A conversational assistant is an LLM specifically tuned to be helpful in back-and-forth dialogue — to follow your instructions, ask for clarification, refuse harmful requests, and stay on task across several turns. It is the polished, cooperative personality you meet in products like ChatGPT or Claude, not the raw model underneath.

That cooperative behavior is trained in after the base model learns language. Humans, or other models, show it examples of good replies and rank competing answers, and the model is adjusted to prefer the helpful, honest, harmless ones. This stage, often called instruction tuning and alignment, is what turns a text predictor into something that feels like an assistant.