VOCABULARY
Active Learning
Active learning is a special case of machine learning in which a learning algorithm can interactively query a user to label new data points with the desired outputs.
How Active Learning Works
Initialization:
- Start with a small labeled dataset and train an initial model.
- Have a much larger pool of unlabeled data.
Uncertainty Sampling:
- The model makes predictions on the unlabeled data.
- It then identifies instances where it's most uncertain (e.g., those for which the predicted class probabilities are closest to 50% in a binary classification task).
Query for Labels:
- The model (or system) queries the user or expert to label the instances it's uncertain about.
Update the Model:
- Incorporate the newly labeled instances into the training dataset.
- Retrain the model using the updated dataset.
Iterate:
- The process is repeated, with the model iteratively selecting uncertain data points, querying for their labels, and updating its knowledge.
![Lakera LLM Security Playbook](https://cdn.prod.website-files.com/65080baa3f9a607985451de3/65254b087d7b8af4624a0982_CTA%20Image.webp)
Learn how to protect against the most common LLM vulnerabilities
Download this guide to delve into the most common LLM security risks and ways to mitigate them.
Related terms
![](https://cdn.prod.website-files.com/65080baa3f9a607985451de3/650d8986cacb870bc87f91f2_Spacer%20Bottomr.webp)
Activate
untouchable mode.
untouchable mode.
Get started for free.
Lakera Guard protects your LLM applications from cybersecurity risks with a single line of code. Get started in minutes. Become stronger every day.
Join our Slack Community.
Several people are typing about AI/ML security. 
Come join us and 1000+ others in a chat that’s thoroughly SFW.