Instruction Tuning
Continued training on instruction-response data so a model follows human instructions better.
In plain language
A pretrained model only continues text; instruction tuning turns it into an assistant. Training data contains task descriptions, inputs and ideal responses across QA, summarization, writing and refusal. It improves instruction-following and format stability, not knowledge injection; factual updates still require retrieval or continued pretraining. Data quality matters more than volume: a small, diverse, high-quality set often beats mass-produced templates.
Example
After instruction tuning, the same base model follows “answer in three bullets” instead of continuing with prose.
Recommended tools
Related terms
Alignment
Research and engineering to make AI systems behave according to human intent, values and safety constraints.
Fine-tuning
Continue training a model on task data to better match a style or task.
Reinforcement Learning from Human Feedback
Train a reward model from human preferences, then optimize model outputs with reinforcement learning.
Agent Memory
A mechanism for an agent to store and retrieve user preferences, past tasks and key facts across sessions.