From Allies to Adversaries: Manipulating LLM Tool-Calling through Adversarial Injection
Paper
•
2412.10198
•
Published
Trevor Zhang
zrpxx
·
AI & ML interests
None yet
Recent Activity
authored
a paper
about 2 months ago
From Allies to Adversaries: Manipulating LLM Tool-Calling through
Adversarial Injection
authored
a paper
about 2 months ago
AdInject: Real-World Black-Box Attacks on Web Agents via Advertising
Delivery
authored
a paper
about 2 months ago
Joint-GCG: Unified Gradient-Based Poisoning Attacks on
Retrieval-Augmented Generation Systems
Organizations
