Google TechTalks
Atomic Facts to Structured Knowledge: Rethinking Unlearning & Jailbreaking in Large Language Models
This talk reveals how the interconnected nature of knowledge within Large Language Models creates fundamental vulnerabilities, enabling sophisticated jailbreaking attacks and undermining current unlearning methods.
Machine Text Detectors are Membership Inference Attacks
This research reveals that machine text detection and membership inference attacks, traditionally studied as separate problems, are fundamentally linked both theoretically and empirically, sharing optimal methods and exhibiting high cross-task transferability.
Cascading Adversarial Bias from Injection to Distillation in Language Models
Adversarial bias injected into large language models (LLMs) during instruction tuning can cascade and amplify in distilled student models, even with minimal poisoning, bypassing current detection methods.
Persistent Pre-Training Poisoning of LLMs
Adversaries can persistently compromise Large Language Models (LLMs) by injecting a small amount of malicious data (as little as 10 tokens per million) into their pre-training datasets, leading to behaviors like denial of service, private data extraction, and belief manipulation, even after subsequent alignment training.
Want more on ai safety?
Explore deep-dive summaries and actionable takeaways from the best minds across different podcasts discussing this topic.
View All Ai Safety Episodes→Don't see the episode you're looking for?
We're constantly adding new episodes, but if you want to see a specific one from Google TechTalks summarized, let us know!
Submit an Episode