4 papers
Large Vision-Language Models Get Lost in Attention
Gongli Xi, Ye Tian, Mengyu Yang +5
Despite the rapid evolution of training paradigms, the decoder backbone of large vision--language models (LVLMs) remains fundamentally rooted in the residual-connection Transformer…
Invariant Causal Routing for Governing Social Norms in Online Market Economies
Xiangning Yu, Qirui Mi, Xiao Xue +4
Social norms are stable behavioral patterns that emerge endogenously within economic systems through repeated interactions among agents. In online market economies, such norms -- l…
CoPHo: Classifier-guided Conditional Topology Generation with Persistent Homology
Gongli Xi, Ye Tian, Mengyu Yang +5
The structure of topology underpins much of the research on performance and robustness, yet available topology data are typically scarce, necessitating the generation of synthetic…
AdaViPro: Region-based Adaptive Visual Prompt for Large-Scale Models Adapting
Mengyu Yang, Ye Tian, Lanshan Zhang +3
Recently, prompt-based methods have emerged as a new alternative `parameter-efficient fine-tuning' paradigm, which only fine-tunes a small number of additional parameters while kee…