8 papers
SodaMem: Evidence-Grounded Temporal Graph Memory for LLM Agents
Fengrong Wan, Chengcan Wu, Ningtao Lyu
Large language model (LLM) agents that assist users over weeks of conversation must remember what is currently true, not merely what was once said. Flat RAG diaries and Markdown lo…
Absorber LLM: Harnessing Causal Synchronization for Test-Time Training
Zhixin Zhang, Shabo Zhang, Chengcan Wu +2
Transformers suffer from a high computational cost that grows with sequence length for self-attention, making inference in long streams prohibited by memory consumption. Constant-m…
AgentWorm: Self-Propagating Attacks Across LLM Agent Ecosystems
Yihao Zhang, Zeming Wei, Xiaokun Luan +7
Autonomous LLM-based agents increasingly operate as long-running processes forming densely interconnected multi-agent ecosystems, whose security properties remain largely unexplore…
RACC: Representation-Aware Coverage Criteria for LLM Safety Testing
Zeming Wei, Zhixin Zhang, Chengcan Wu +3
Large Language Models (LLMs) face severe safety risks from jailbreak attacks, yet current safety testing largely relies on static datasets and lacks systematic criteria to evaluate…
Securing Multi-Agent Systems Against Corruptions via Node Contribution Backpropagation
Chengcan Wu, Zhixin Zhang, Mingqian Xu +2
Multi-Agent Systems (MAS) have become a prevalent paradigm for Large Language Model (LLM) applications. However, the complex multi-agent design in MAS introduces unique trustworthi…
Reliable Unlearning Harmful Information in LLMs with Metamorphosis Representation Projection
Chengcan Wu, Zeming Wei, Huanran Chen +2
While Large Language Models (LLMs) have demonstrated impressive performance in various domains and tasks, concerns about their safety are becoming increasingly severe. In particula…
ReGA: Model-Based Safeguard for LLMs via Representation-Guided Abstraction
Zeming Wei, Chengcan Wu, Meng Sun
Large Language Models (LLMs) have achieved tremendous success in various tasks, yet concerns about their safety and security have emerged. In particular, they pose risks of generat…
Secure LLM Fine-Tuning via Safety-Aware Probing
Chengcan Wu, Zhixin Zhang, Zeming Wei +3
Large language models (LLMs) have achieved remarkable success across many applications, but their ability to generate harmful content raises serious safety concerns. Although safet…