16 papers
TIGA: Trajectory-Injected Generative Attack against Black-box AIGC Detectors
Xia Du, Zhuosen Bao, Zheng Lin +6
Recent diffusion models have achieved remarkable realism in facial image synthesis, posing growing challenges to artificial intelligence-generated content (AIGC) forensic detectors…
PACT: Privileged Trace Co-Training for Multi-Turn Tool-Use Agents
Zhenbang Du, Jun Luo, Zhiwei Zheng +8
Multi-turn tool-use agents must reason, call tools, and adapt to observations across several interaction turns. Post-training such agents is challenging, as reinforcement learning…
Conflict-Aware Federated Fine-Tuning of Large Language Models with Mixture-of-Experts
Yijun Lu, Zihan Fang, Pengpeng Qiao +6
The continuous scaling of large language models (LLMs) incurs prohibitive computational costs, making Mixture-of-Experts (MoE) a scalable alternative for efficient fine-tuning via…
What Makes LVLMs Hallucinate Less? Unveiling the Architectural Factors Behind Hallucination Robustness
Yusheng He, Jizhe Zhou, Xia Du +3
Hallucination remains one of the key challenges undermining the reliability of Large Vision-Language Models (LVLMs). But what makes an LVLM hallucinate less? Many existing efforts…
FluxShard: Motion-Aware Feature Cache Reuse for Collaborative Video Analytics in Mobile Edge Computing
Xiuxian Guan, Zongyuan Zhang, Zheng Lin +8
Caching and reusing intermediate features across consecutive frames is a common technique to reduce redundant computation and transmission for edge-cloud video analytics in mobile…
Physically-Induced Atmospheric Adversarial Perturbations: Enhancing Transferability and Robustness in Remote Sensing Image Classification
Weiwei Zhuang, Wangze Xie, Qi Zhang +9
Adversarial attacks pose a severe threat to the reliability of deep learning models in remote sensing (RS) image classification. Most existing methods rely on direct pixel-wise per…