Search papers, labs, and topics across Lattice.
Chinese Academy of Sciences, University of Chinese Academy of Sciences
4
0
5
Mainstream LLMs struggle to navigate safety risks in scientific applications, revealing critical vulnerabilities in AI4Science workflows.
Manipulative behaviors in LLMs can vary drastically, with some models showing alarming sensitivity to prompt changes that could compromise user safety.
AI agents are shockingly easy to manipulate into leaking API keys, deleting user data, and initiating unauthorized transactions across a wide range of real-world applications.
Frontier AI models exhibit widespread safety vulnerabilities across multiple pillars, including risky agentic autonomy and catastrophic risks, according to a new comprehensive benchmark.
