Deliberately corrupting training data to manipulate a model's behavior.
Deliberately corrupting training data to manipulate a model's behavior. An attacker could insert carefully crafted examples that cause the model to behave incorrectly on specific inputs. A growing security concern as models are trained on internet-scale data that's hard to fully audit.
The broad field studying how to build AI systems that are safe, reliable, and beneficial.
A mathematical function applied to a neuron's output that introduces non-linearity into the network.
An optimization algorithm that combines the best parts of two other methods — AdaGrad and RMSProp.
Artificial General Intelligence.
The research field focused on making sure AI systems do what humans actually want them to do.
The science of creating machines that can perform tasks requiring human-like intelligence — reasoning, learning, perception, language understanding, and decision-making.
Browse our complete glossary or subscribe to our newsletter for the latest AI news and insights.