14 papers
Beyond Episodic Evaluation: Memory Architectural Bottlenecks in Sequential Embodied Question Answering
Zikui Cai, Kaushal Janga, Tan Dat Dao +15
Embodied question answering (EQA) is traditionally evaluated under an episodic formulation, where agents solve each task independently and reset internal state between episodes. Ho…
DecodingTrust-Agent Platform (DTap): A Controllable and Interactive Red-Teaming Platform for AI Agents
Zhaorun Chen, Xun Liu, Haibo Tong +14
AI agents are increasingly deployed across diverse domains to automate complex workflows through long-horizon and high-stakes action executions. Due to their high capability and fl…
AudioGuard: Toward Comprehensive Audio Safety Protection Across Diverse Threat Models
Mintong Kang, Chen Fang, Bo Li
Audio has rapidly become a primary interface for foundation models, powering real-time voice assistants. Ensuring safety in audio systems is inherently more complex than just "unsa…
Poly-Guard: Massive Multi-Domain Safety Policy-Grounded Guardrail Dataset
Mintong Kang, Zhaorun Chen, Chejian Xu +6
As LLMs become widespread across diverse applications, concerns about the security and safety of LLM interactions have intensified. Numerous guardrail models and benchmarks have be…
FG-OrIU: Towards Better Forgetting via Feature-Gradient Orthogonality for Incremental Unlearning
Qian Feng, JiaHang Tu, Mintong Kang +3
Incremental unlearning (IU) is critical for pre-trained models to comply with sequential data deletion requests, yet existing methods primarily suppress parameters or confuse knowl…
Mitigating Indirect Prompt Injection via Instruction-Following Intent Analysis
Mintong Kang, Chong Xiang, Sanjay Kariyappa +3
Indirect prompt injection attacks (IPIAs), where large language models (LLMs) follow malicious instructions hidden in input data, pose a critical threat to LLM-powered agents. In t…