Zobrazeno 1 - 10
of 44 928
pro vyhledávání: '"P. A. Bowen"'
Recent advances in spin transport research have highlighted the potential of quantum paramagnets as platforms for exploring novel phenomena and developing next-generation technologies. In this paper, we investigate the flavor Nernst effect (FNE) in q
Externí odkaz:
http://arxiv.org/abs/2408.15160
Length extrapolation algorithms based on Rotary position embedding (RoPE) have shown promising results in extending the context length of language models. However, understanding how position embedding can capture longer-range contextual information r
Externí odkaz:
http://arxiv.org/abs/2410.08703
Autor:
Li, Yadong, Sun, Haoze, Lin, Mingan, Li, Tianpeng, Dong, Guosheng, Zhang, Tao, Ding, Bowen, Song, Wei, Cheng, Zhenglin, Huo, Yuqi, Chen, Song, Li, Xu, Pan, Da, Zhang, Shusen, Wu, Xin, Liang, Zheng, Liu, Jun, Lu, Keer, Zhao, Yaqi, Shen, Yanjun, Yang, Fan, Yu, Kaicheng, Lin, Tao, Xu, Jianhua, Zhou, Zenan, Chen, Weipeng
The salient multimodal capabilities and interactive experience of GPT-4o highlight its critical role in practical applications, yet it lacks a high-performing open-source counterpart. In this paper, we introduce Baichuan-Omni, the first open-source 7
Externí odkaz:
http://arxiv.org/abs/2410.08565
Ruling a wide range of phenomena, dipole-dipole interactions (DDI) are typically constrained to the short range due to their rapid decay with the increasing dipole separations, limiting the performance in long-range applications. By judiciously desig
Externí odkaz:
http://arxiv.org/abs/2410.07907
While advancements in NLP have significantly improved the performance of Large Language Models (LLMs) on tasks requiring vertical thinking, their lateral thinking capabilities remain under-explored and challenging to measure due to the complexity of
Externí odkaz:
http://arxiv.org/abs/2410.06733
In various scientific and engineering fields, the primary research areas have revolved around physics-based dynamical systems modeling and data-driven time series analysis. According to the embedding theory, dynamical systems and time series can be m
Externí odkaz:
http://arxiv.org/abs/2410.06651
Autor:
Ye, Hanrong, Zhang, Haotian, Daxberger, Erik, Chen, Lin, Lin, Zongyu, Li, Yanghao, Zhang, Bowen, You, Haoxuan, Xu, Dan, Gan, Zhe, Lu, Jiasen, Yang, Yinfei
This research aims to comprehensively explore building a multimodal foundation model for egocentric video understanding. To achieve this goal, we work on three fronts. First, as there is a lack of QA data for egocentric video understanding, we develo
Externí odkaz:
http://arxiv.org/abs/2410.07177
Publikováno v:
NeurIPs 2024
In this paper, we approach an overlooked yet critical task Graph2Image: generating images from multimodal attributed graphs (MMAGs). This task poses significant challenges due to the explosion in graph size, dependencies among graph entities, and the
Externí odkaz:
http://arxiv.org/abs/2410.07157
Neural Green's Function Accelerated Iterative Methods for Solving Indefinite Boundary Value Problems
Neural operators, which learn mappings between the function spaces, have been applied to solve boundary value problems in various ways, including learning mappings from the space of the forcing terms to the space of the solutions with the substantial
Externí odkaz:
http://arxiv.org/abs/2410.06919
The soft-argmax operation is widely adopted in neural network-based stereo matching methods to enable differentiable regression of disparity. However, network trained with soft-argmax is prone to being multimodal due to absence of explicit constraint
Externí odkaz:
http://arxiv.org/abs/2410.06527