Let G be a special linear group over the real,the complex or the quaternion,or a special unitary group.In this note,we determine all special unipotent representations of G in the sense of Arthur and Barbasch-Vogan,and...Let G be a special linear group over the real,the complex or the quaternion,or a special unitary group.In this note,we determine all special unipotent representations of G in the sense of Arthur and Barbasch-Vogan,and show in particular that all of them are unitarizable.展开更多
We study the transfer between small special unipotent representations for all equal rank real forms of type E_(6) and E_(7). As a consequence, one can verify these modules are unitarity using the results of Wallach an...We study the transfer between small special unipotent representations for all equal rank real forms of type E_(6) and E_(7). As a consequence, one can verify these modules are unitarity using the results of Wallach and Zhu. Moreover, the K-spectra of these modules can be obtained explicitly.展开更多
首先利用bidirectional encoder representations from transformers(BERT)模型的强大的语境理解能力来提取数据法律文本的深层语义特征,然后引入细粒度特征提取层,依照注意力机制,重点关注文本中与数据法律问答相关的关键部分,最后对...首先利用bidirectional encoder representations from transformers(BERT)模型的强大的语境理解能力来提取数据法律文本的深层语义特征,然后引入细粒度特征提取层,依照注意力机制,重点关注文本中与数据法律问答相关的关键部分,最后对所采集的法律问答数据集进行训练和评估.结果显示:与传统的多个单一模型相比,所提出的模型在准确度、精确度、召回率、F1分数等关键性能指标上均有提升,表明该系统能够更有效地理解和回应复杂的数据法学问题,为研究数据法学的专业人士和公众用户提供更高质量的问答服务.展开更多
Learning modality-fused representations and processing unaligned multimodal sequences are meaningful and challenging in multimodal emotion recognition.Existing approaches use directional pairwise attention or a messag...Learning modality-fused representations and processing unaligned multimodal sequences are meaningful and challenging in multimodal emotion recognition.Existing approaches use directional pairwise attention or a message hub to fuse language,visual,and audio modalities.However,these fusion methods are often quadratic in complexity with respect to the modal sequence length,bring redundant information and are not efficient.In this paper,we propose an efficient neural network to learn modality-fused representations with CB-Transformer(LMR-CBT)for multimodal emotion recognition from unaligned multi-modal sequences.Specifically,we first perform feature extraction for the three modalities respectively to obtain the local structure of the sequences.Then,we design an innovative asymmetric transformer with cross-modal blocks(CB-Transformer)that enables complementary learning of different modalities,mainly divided into local temporal learning,cross-modal feature fusion and global self-attention representations.In addition,we splice the fused features with the original features to classify the emotions of the sequences.Finally,we conduct word-aligned and unaligned experiments on three challenging datasets,IEMOCAP,CMU-MOSI,and CMU-MOSEI.The experimental results show the superiority and efficiency of our proposed method in both settings.Compared with the mainstream methods,our approach reaches the state-of-the-art with a minimum number of parameters.展开更多
This study explores the ground-state characteristics of neutron-rich sodium isotopes,encompassing two-neutron separation energies,root-mean-square radii,quadrupole moments of proton and neutron distributions,single-pa...This study explores the ground-state characteristics of neutron-rich sodium isotopes,encompassing two-neutron separation energies,root-mean-square radii,quadrupole moments of proton and neutron distributions,single-particle levels of bound and resonant states,and neutron density distributions and shapes.Simultaneously,special attention is paid to the distinctive physical phenomena associated with these isotopes.The deformed relativistic mean field theory in complex momentum representations with BCS pairings(DRMF-CMR-BCS)employed in our research provides resonant states with real physics,offering insights into deformed halo nuclei.Four effective interactions(NL3,NL3^(*),PK1,and NLSH)were considered to assess the influence of continuum and deformation effects on halo structures.Calculations for odd-even nuclei ^(35–43)Na revealed the dependence on the chosen effective interaction and number of considered resonant states.Neutron occupation patterns near the Fermi surface,particularly in orbitals 1/2^(−)_(3) and 3/2^(−)_(2),were determined to be crucial in halo formation.The study provided detailed insights into the density distributions,shape evolution,and structure of neutron-rich sodium isotopes,contributing valuably to the field of nuclear physics.展开更多
We prove a converse theorem for split even special orthogonal groups over finite fields.This is the only case left on converse theorems of classical groups and the difficulty is the existence of the outer automorphism...We prove a converse theorem for split even special orthogonal groups over finite fields.This is the only case left on converse theorems of classical groups and the difficulty is the existence of the outer automorphism. In this paper, we develop new ideas and overcome this difficulty.展开更多
The rapid expansion of online content and big data has precipitated an urgent need for efficient summarization techniques to swiftly comprehend vast textual documents without compromising their original integrity.Curr...The rapid expansion of online content and big data has precipitated an urgent need for efficient summarization techniques to swiftly comprehend vast textual documents without compromising their original integrity.Current approaches in Extractive Text Summarization(ETS)leverage the modeling of inter-sentence relationships,a task of paramount importance in producing coherent summaries.This study introduces an innovative model that integrates Graph Attention Networks(GATs)with Transformer-based Bidirectional Encoder Representa-tions from Transformers(BERT)and Latent Dirichlet Allocation(LDA),further enhanced by Term Frequency-Inverse Document Frequency(TF-IDF)values,to improve sentence selection by capturing comprehensive topical information.Our approach constructs a graph with nodes representing sentences,words,and topics,thereby elevating the interconnectivity and enabling a more refined understanding of text structures.This model is stretched to Multi-Document Summarization(MDS)from Single-Document Summarization,offering significant improvements over existing models such as THGS-GMM and Topic-GraphSum,as demonstrated by empirical evaluations on benchmark news datasets like Cable News Network(CNN)/Daily Mail(DM)and Multi-News.The results consistently demonstrate superior performance,showcasing the model’s robustness in handling complex summarization tasks across single and multi-document contexts.This research not only advances the integration of BERT and LDA within a GATs but also emphasizes our model’s capacity to effectively manage global information and adapt to diverse summarization challenges.展开更多
The structure of the irreducible collective spaces of the group Sp(12,R),which many-particle nuclear states are classified according to the chain Sp(12,R)⊃U(6)⊃SO(6)⊃SU_(pn)(3)⊗SO(2)⊃SO(3)of the proton-neutron symplec...The structure of the irreducible collective spaces of the group Sp(12,R),which many-particle nuclear states are classified according to the chain Sp(12,R)⊃U(6)⊃SO(6)⊃SU_(pn)(3)⊗SO(2)⊃SO(3)of the proton-neutron symplectic model(PNSM),is considered in detail.This chain of the PNSM was recently shown to correspond to a microscopic shell-model version of the Bohr-Mottelson collective model.The construction of the relevant shell-model representations of the Sp(12,R)group along this chain is considered for three nuclei with varying collective properties and from different mass regions.It is shown that the SU_(pn)(3)basis states of the Sp(12,R)representations belonging to the SO(6)irreps with seniorityυ≥υ0,withυ0 denoting the maximal seniority SO(6)irrep contained in the Sp(12,R)bandhead,are always Pauli allowed,but organized in a different way into different SO(6)shells.This is in contrast to the case of filling the levels of the standard three-dimensional harmonic oscillator and using the plethysm operation.Although the SU_(pn)(3)multiplets withinυ<υ0 are not all Pauli forbidden,it is safe to discard them.The results obtained in the present work are important for the practical application of the microscopic version of the Bohr-Mottelson collective model.展开更多
基金supported by NSF grant (Award Number 2000254)supported by the National Natural Science Foundation of China (Grant Nos. 11701364 and 11971305)+4 种基金Xiamen University Malaysia Research Fund (Grant No. XMUMRF/2022-C9/IMAT/0019)supported by National Key R&D Program of China (Grant Nos. 2022YFA1005300 and 2020YFA0712600)New Cornerstone Investigator Programsupported by MOE AcRF Tier 1 grant A-0004280-00-00Provost’s Chair grant E-146-000-052-001 in NUS
文摘Let G be a special linear group over the real,the complex or the quaternion,or a special unitary group.In this note,we determine all special unipotent representations of G in the sense of Arthur and Barbasch-Vogan,and show in particular that all of them are unitarizable.
基金supported by Natural Science Foundation of Shanghai (Grant No. 22ZR1422900)supported by the National Natural Science Foundation of China (Grant No. 11271460)+3 种基金Guangdong Province(Grant No. 2023A1515012186)Shenzhen City (Grant No. 2022373357)the Research Grants Council of HKSAR,China (Grant No. 16302521)supported by Shenzhen Science and Technology Innovation Committee (Grant No. 20220818094918001)
文摘We study the transfer between small special unipotent representations for all equal rank real forms of type E_(6) and E_(7). As a consequence, one can verify these modules are unitarity using the results of Wallach and Zhu. Moreover, the K-spectra of these modules can be obtained explicitly.
文摘首先利用bidirectional encoder representations from transformers(BERT)模型的强大的语境理解能力来提取数据法律文本的深层语义特征,然后引入细粒度特征提取层,依照注意力机制,重点关注文本中与数据法律问答相关的关键部分,最后对所采集的法律问答数据集进行训练和评估.结果显示:与传统的多个单一模型相比,所提出的模型在准确度、精确度、召回率、F1分数等关键性能指标上均有提升,表明该系统能够更有效地理解和回应复杂的数据法学问题,为研究数据法学的专业人士和公众用户提供更高质量的问答服务.
基金National Natural Science Foundation of China(Grant No.72293583).
文摘Learning modality-fused representations and processing unaligned multimodal sequences are meaningful and challenging in multimodal emotion recognition.Existing approaches use directional pairwise attention or a message hub to fuse language,visual,and audio modalities.However,these fusion methods are often quadratic in complexity with respect to the modal sequence length,bring redundant information and are not efficient.In this paper,we propose an efficient neural network to learn modality-fused representations with CB-Transformer(LMR-CBT)for multimodal emotion recognition from unaligned multi-modal sequences.Specifically,we first perform feature extraction for the three modalities respectively to obtain the local structure of the sequences.Then,we design an innovative asymmetric transformer with cross-modal blocks(CB-Transformer)that enables complementary learning of different modalities,mainly divided into local temporal learning,cross-modal feature fusion and global self-attention representations.In addition,we splice the fused features with the original features to classify the emotions of the sequences.Finally,we conduct word-aligned and unaligned experiments on three challenging datasets,IEMOCAP,CMU-MOSI,and CMU-MOSEI.The experimental results show the superiority and efficiency of our proposed method in both settings.Compared with the mainstream methods,our approach reaches the state-of-the-art with a minimum number of parameters.
基金Partly supported by the National Natural Science Foundation of China (11935001, 11575001)the Natural Science Foundation of Anhui Province(2008085MA26)+2 种基金Anhui project (Z010118169)the Heavy Ion Research Facility in Lanzhou (HIRFL)(HIR2021PY007)the project of Key Laboratory of High Precision Nuclear Spectroscopy conducted in Chinese Academy of Sciences.
文摘This study explores the ground-state characteristics of neutron-rich sodium isotopes,encompassing two-neutron separation energies,root-mean-square radii,quadrupole moments of proton and neutron distributions,single-particle levels of bound and resonant states,and neutron density distributions and shapes.Simultaneously,special attention is paid to the distinctive physical phenomena associated with these isotopes.The deformed relativistic mean field theory in complex momentum representations with BCS pairings(DRMF-CMR-BCS)employed in our research provides resonant states with real physics,offering insights into deformed halo nuclei.Four effective interactions(NL3,NL3^(*),PK1,and NLSH)were considered to assess the influence of continuum and deformation effects on halo structures.Calculations for odd-even nuclei ^(35–43)Na revealed the dependence on the chosen effective interaction and number of considered resonant states.Neutron occupation patterns near the Fermi surface,particularly in orbitals 1/2^(−)_(3) and 3/2^(−)_(2),were determined to be crucial in halo formation.The study provided detailed insights into the density distributions,shape evolution,and structure of neutron-rich sodium isotopes,contributing valuably to the field of nuclear physics.
基金partially supported by the NSF Grants DMS-1848058partially supported by the NSF Grants DMS-1702218, DMS-1848058start-up funds from the Department of Mathematics at Purdue University.
文摘We prove a converse theorem for split even special orthogonal groups over finite fields.This is the only case left on converse theorems of classical groups and the difficulty is the existence of the outer automorphism. In this paper, we develop new ideas and overcome this difficulty.
文摘The rapid expansion of online content and big data has precipitated an urgent need for efficient summarization techniques to swiftly comprehend vast textual documents without compromising their original integrity.Current approaches in Extractive Text Summarization(ETS)leverage the modeling of inter-sentence relationships,a task of paramount importance in producing coherent summaries.This study introduces an innovative model that integrates Graph Attention Networks(GATs)with Transformer-based Bidirectional Encoder Representa-tions from Transformers(BERT)and Latent Dirichlet Allocation(LDA),further enhanced by Term Frequency-Inverse Document Frequency(TF-IDF)values,to improve sentence selection by capturing comprehensive topical information.Our approach constructs a graph with nodes representing sentences,words,and topics,thereby elevating the interconnectivity and enabling a more refined understanding of text structures.This model is stretched to Multi-Document Summarization(MDS)from Single-Document Summarization,offering significant improvements over existing models such as THGS-GMM and Topic-GraphSum,as demonstrated by empirical evaluations on benchmark news datasets like Cable News Network(CNN)/Daily Mail(DM)and Multi-News.The results consistently demonstrate superior performance,showcasing the model’s robustness in handling complex summarization tasks across single and multi-document contexts.This research not only advances the integration of BERT and LDA within a GATs but also emphasizes our model’s capacity to effectively manage global information and adapt to diverse summarization challenges.
文摘The structure of the irreducible collective spaces of the group Sp(12,R),which many-particle nuclear states are classified according to the chain Sp(12,R)⊃U(6)⊃SO(6)⊃SU_(pn)(3)⊗SO(2)⊃SO(3)of the proton-neutron symplectic model(PNSM),is considered in detail.This chain of the PNSM was recently shown to correspond to a microscopic shell-model version of the Bohr-Mottelson collective model.The construction of the relevant shell-model representations of the Sp(12,R)group along this chain is considered for three nuclei with varying collective properties and from different mass regions.It is shown that the SU_(pn)(3)basis states of the Sp(12,R)representations belonging to the SO(6)irreps with seniorityυ≥υ0,withυ0 denoting the maximal seniority SO(6)irrep contained in the Sp(12,R)bandhead,are always Pauli allowed,but organized in a different way into different SO(6)shells.This is in contrast to the case of filling the levels of the standard three-dimensional harmonic oscillator and using the plethysm operation.Although the SU_(pn)(3)multiplets withinυ<υ0 are not all Pauli forbidden,it is safe to discard them.The results obtained in the present work are important for the practical application of the microscopic version of the Bohr-Mottelson collective model.
