Defending Against Indirect Prompt Injection Attacks With Spotlighting
25:39
Let’s Make it Personal: Customizing Threat Intelligence with Metric Learning
29:34
End-to-End Framework using LLMs for Technique Identification and Threat-Actor Attribution
23:32
Defending Large Language Models Against Attacks With Residual Stream Activation Analysis
24:02
A Corrida para Aproveitar o Poder Alucinante da Computação Quântica | The Future com Hannah Fry
19:12
Hamm-Grams: Mining Common Regular Expressions via Locality Sensitive Hashing
20:33
PyRIT: A Framework for Security Risk Identification and Red Teaming in Generative AI Systems
29:11
AdapterSwap: Continuous Training of LLMs with Data Removal and Access-Control Guarantees
26:07