Summary of Cross-task Attack: a Self-supervision Generative Framework Based on Attention Shift, by Qingyuan Zeng et al.
Cross-Task Attack: A Self-Supervision Generative Framework Based on Attention Shiftby Qingyuan Zeng, Yunpeng Gong, Min…
Cross-Task Attack: A Self-Supervision Generative Framework Based on Attention Shiftby Qingyuan Zeng, Yunpeng Gong, Min…
LLMs as Function Approximators: Terminology, Taxonomy, and Questions for Evaluationby David SchlangenFirst submitted to arxiv…
Black-Box Opinion Manipulation Attacks to Retrieval-Augmented Generation of Large Language Modelsby Zhuo Chen, Jiawei Liu,…
Latent Causal Probing: A Formal Perspective on Probing with Causal Models of Databy Charles Jin,…
Addressing Imbalance for Class Incremental Learning in Medical Image Classificationby Xuze Hao, Wenqian Ni, Xuhao…
Phi-3 Safety Post-Training: Aligning Language Models with a “Break-Fix” Cycleby Emman Haider, Daniel Perez-Becker, Thomas…
High Risk of Political Bias in Black Box Emotion Inference Modelsby Hubert Plisiecki, Paweł Lenartowicz,…
LinSATNet: The Positive Linear Satisfiability Neural Networksby Runzhong Wang, Yunhao Zhang, Ziao Guo, Tianyi Chen,…
DuoFormer: Leveraging Hierarchical Visual Representations by Local and Global Attentionby Xiaoya Tang, Bodong Zhang, Beatrice…
RT-Pose: A 4D Radar Tensor-based 3D Human Pose Estimation and Localization Benchmarkby Yuan-Hao Ho, Jen-Hao…