Zobrazeno 1 - 10
of 1 288
pro vyhledávání: '"Qiu, Han"'
Hallucination, a phenomenon where multimodal large language models~(MLLMs) tend to generate textual responses that are plausible but unaligned with the image, has become one major hurdle in various MLLM-related applications. Several benchmarks have b
Externí odkaz:
http://arxiv.org/abs/2410.09962
With the rapidly increasing number of satellites in space and their enhanced capabilities, the amount of earth observation images collected by satellites is exceeding the transmission limits of satellite-to-ground links. Although existing learned ima
Externí odkaz:
http://arxiv.org/abs/2410.01698
Autor:
Xu, Rongwu, Cai, Yishuo, Zhou, Zhenhong, Gu, Renjie, Weng, Haiqin, Liu, Yan, Zhang, Tianwei, Xu, Wei, Qiu, Han
The risk of harmful content generated by large language models (LLMs) becomes a critical concern. This paper presents a systematic study on assessing and improving LLMs' capability to perform the task of \textbf{course-correction}, \ie, the model can
Externí odkaz:
http://arxiv.org/abs/2407.16637
The common toxicity and societal bias in contents generated by large language models (LLMs) necessitate strategies to reduce harm. Present solutions often demand white-box access to the model or substantial training, which is impractical for cutting-
Externí odkaz:
http://arxiv.org/abs/2407.15366
We are witnessing a surge in the use of commercial off-the-shelf (COTS) hardware for cost-effective in-orbit computing, such as deep neural network (DNN) based on-satellite sensor data processing, Earth object detection, and task decision.However, on
Externí odkaz:
http://arxiv.org/abs/2407.11853
Inspired by the success of general-purpose models in NLP, recent studies attempt to unify different vision tasks in the same sequence format and employ autoregressive Transformers for sequence prediction. They apply uni-directional attention to captu
Externí odkaz:
http://arxiv.org/abs/2403.07692
Binary code similarity detection (BCSD) is a fundamental technique for various application. Many BCSD solutions have been proposed recently, which mostly are embedding-based, but have shown limited accuracy and efficiency especially when the volume o
Externí odkaz:
http://arxiv.org/abs/2402.18818
Autor:
Wang, Hao, Gao, Zeyu, Zhang, Chao, Sha, Zihan, Sun, Mingyang, Zhou, Yuchen, Zhu, Wenyu, Sun, Wenju, Qiu, Han, Xiao, Xi
Binary code representation learning has shown significant performance in binary analysis tasks. But existing solutions often have poor transferability, particularly in few-shot and zero-shot scenarios where few or no training samples are available fo
Externí odkaz:
http://arxiv.org/abs/2402.16928
Segment Anything Models (SAMs) like SEEM and SAM have demonstrated great potential in learning to segment anything. The core design of SAMs lies with Promptable Segmentation, which takes a handcrafted prompt as input and returns the expected segmenta
Externí odkaz:
http://arxiv.org/abs/2401.04651
Traditional computer vision generally solves each single task independently by a dedicated model with the task instruction implicitly designed in the model architecture, arising two limitations: (1) it leads to task-specific models, which require mul
Externí odkaz:
http://arxiv.org/abs/2312.16602