Explore Model Thresholds

Thresholds allow security teams to filter and assess which vulnerabilities are most critical to remediate. Organizations have to make tough calls when choosing which vulnerabilities to prioritize, and thresholds allows teams to make educated decisions based off global model data.

Today we added a new feature to our web UI that allows users to explore their scoring thresholds. From the models page, click on the new “Explore Thresholds” tab to check it out. Move the sliders to tune our models by score, coverage percentage, and efficiency percentage, and then click “Run Query” to see CVEs that match your selected threshold.

Coverage is the proportion of vulnerabilities with known exploitation activity that are prioritized. Efficiency is the proportion of all prioritized vulnerabilities with known exploitation activity. We measure our model performance against our dataset of exploitation activity, which as of this blog’s publication covers over 16,000 CVEs, and is updated hourly.

We’re actively exploring different ways to visualize and tune threshold data. If you’re interested in tracking thresholds for your own data, we offer custom built local models. See how your model would differ — book a walkthroughLink to https://www.empiricalsecurity.com/contact.