arXiv subjects

Jun Zhan

Publications and source records attributed to Jun Zhan.

34 records · Page 2Linked to original sources

arXiv2026

RIS-Enhanced Information-Decoupled Symbiotic Radio Over Broadcasting Signals

This paper studies a reconfigurable intelligent surface (RIS)-enhanced decoupled symbiotic radio (SR) system in which a primary transmitter delivers common data to multiple primary receivers (PRs), while a RIS-based backscatter device sends secondary data to a backscatter receiver (BRx). Unlike conventional SR, the BRx performs energy detection and never decodes the primary signal, thereby removing ambiguity and preventing exposure of the primary payload to unintended receivers. In this paper, we formulate the problem as the minimization of the transmit power subject to a common broadcast rate constraint across all PRs and a bit error rate (BER) constraint at the BRx. The problem is nonconvex due to the unit-modulus RIS constraint and coupled quadratic forms. Leveraging a rate-balanced reformulation and a monotonic BER ratio characterization, we develop a low-complexity penalty-based block coordinate descent algorithm with closed-form updates. Numerical results show fast convergence of the proposed algorithm and reduced power consumption of the considered RIS-enhanced information-decoupled SR system over conventional SR baselines.

eess.SP↗

arXiv2025

Time-Reversal Symmetry Breaking Superconducting State and Collective Modes in Kagome Superconductors

We comprehensively study the unconventional pairing and collective modes in the multiband kagome superconductors AV$_3$Sb$_5$ (A=$\mathrm{K},\mathrm{Cs},\mathrm{Rb}$). By solving gap equations at zero temperature, we identify a transition from normal $s++/s\pm$-wave pairing to time-reversal symmetry (TRS) breaking pairing with a variation of inter-pocket interactions or density of states. This TRS breaking pairing originates from the superconducting phase frustration of different Fermi pockets and can account for experimental TRS breaking signal in kagome superconductors. Moreover, we investigate collective modes, including the Higgs, Leggett, and Bogoloubov-Anderson-Goldstone modes, arising from fluctuations of the amplitude, relative phase, and overall phase of the superconducting order parameters, respectively. Remarkably, due to the presence of multibands, one branch of the Leggett modes becomes nearly massless near the TRS breaking transition, providing a compelling smoking-gun signature of TRS-breaking superconductivity, in clear contrast to TRS-breaking charge orders. Our results elucidate the rich superconducting physics and its associated collective modes in kagome metals, and suggest feasible experimental detection of TRS breaking pairing.

cond-mat.supr-con↗

arXiv2025

Mirror-Selective Quasiparticle Interference in Bilayer Nickelate Superconductor

The recent discovery of high-temperature superconductivity in both bulk and thin-film bilayer nickelates has garnered significant attention. In this study, inspired by recent STM experiments on thin films, we investigate the quasiparticle interference (QPI) characteristics of bilayer nickelates in both normal and superconducting states to identify their Fermiology and pairing symmetry. We demonstrate that the mirror symmetry inherent in the bilayer structure induces mirror-selective quasiparticle scattering by establishing selection rules based on the mirror properties of impurities and the mirror eigenvalues of electronic wavefunctions. This mirror-selective scattering allows for the differentiation of distinct Fermiologies, as QPI patterns vary markedly between scenarios with and without the $d_{z^2}$-bonding Fermi surface (FS). Furthermore, it enables the separate detection of sign changes in superconducting gaps both within the same FS and between different FSs. Crucially, if the mirror-symmetry-enforced selection rules are ignored, the QPI response of an $s_\pm$-wave state can masquerade as that of a conventional $s$-wave state, leading to a misidentification of the pairing symmetry. When combined with field-dependent and reference QPI measurements, this approach facilitates the precise determination of pairing symmetry, even in the presence of FS-dependent gaps and gap anisotropy. Additionally, we discuss practical considerations for STM measurements to effectively identify the pairing symmetry. Our findings demonstrate that mirror-selective QPI is a powerful tool for distinguishing between different Fermiologies and pairing states, which is helpful in pinning down pairing symmetry and revealing the pairing mechanism in bilayer nickelates.

cond-mat.supr-con↗

arXiv2025

Opposite-Mirror-Parity Scattering as the Origin of Superconductivity in Strained Bilayer Nickelates

We study the electronic structure and doping-dependent instabilities of strained La$_3$Ni$_2$O$_7$ thin films using first-principles and functional renormalization group methods. We demonstrate that ordering tendencies are governed by Fermi surface scattering between electrons of opposite mirror parity. Under moderate hole doping, when the $d_{z^2}$ bonding band becomes incipient or crosses the Fermi level, robust $s_{\pm}$-wave superconductivity emerges from cooperative interlayer pairing reinforced by two competing spin-density-wave fluctuations. Compressive strain favors superconductivity in NiO$_2$ bilayers slightly away from the interface, whereas tensile strain induces pair-breaking nesting that suppresses pairing. Our results establish a unified microscopic scenario for superconductivity in pressurized bulk and strained thin-film nickelates, providing new insights into high-T$_c$ pairing in correlated quantum materials.

cond-mat.supr-con↗

arXiv2025

VStyle: A Benchmark for Voice Style Adaptation with Spoken Instructions

Spoken language models (SLMs) have emerged as a unified paradigm for speech understanding and generation, enabling natural human machine interaction. However, while most progress has focused on semantic accuracy and instruction following, the ability of SLMs to adapt their speaking style based on spoken instructions has received limited attention. We introduce Voice Style Adaptation (VSA), a new task that examines whether SLMs can modify their speaking style, such as timbre, prosody, or persona following natural language spoken commands. To study this task, we present VStyle, a bilingual (Chinese & English) benchmark covering four categories of speech generation: acoustic attributes, natural language instruction, role play, and implicit empathy. We also introduce the Large Audio Language Model as a Judge (LALM as a Judge) framework, which progressively evaluates outputs along textual faithfulness, style adherence, and naturalness, ensuring reproducible and objective assessment. Experiments on commercial systems and open source SLMs demonstrate that current models face clear limitations in controllable style adaptation, highlighting both the novelty and challenge of this task. By releasing VStyle and its evaluation toolkit, we aim to provide the community with a foundation for advancing human centered spoken interaction. The dataset and code are publicly available at \href{https://junzhan2000.github.io/VStyle.github.io/}{project's homepage}.

cs.SD↗

arXiv2025

YuE: Scaling Open Foundation Models for Long-Form Music Generation

We tackle the task of long-form music generation--particularly the challenging \textbf{lyrics-to-song} problem--by introducing YuE, a family of open foundation models based on the LLaMA2 architecture. Specifically, YuE scales to trillions of tokens and generates up to five minutes of music while maintaining lyrical alignment, coherent musical structure, and engaging vocal melodies with appropriate accompaniment. It achieves this through (1) track-decoupled next-token prediction to overcome dense mixture signals, (2) structural progressive conditioning for long-context lyrical alignment, and (3) a multitask, multiphase pre-training recipe to converge and generalize. In addition, we redesign the in-context learning technique for music generation, enabling versatile style transfer (e.g., converting Japanese city pop into an English rap while preserving the original accompaniment) and bidirectional generation. Through extensive evaluation, we demonstrate that YuE matches or even surpasses some of the proprietary systems in musicality and vocal agility. In addition, fine-tuning YuE enables additional controls and enhanced support for tail languages. Furthermore, beyond generation, we show that YuE's learned representations can perform well on music understanding tasks, where the results of YuE match or exceed state-of-the-art methods on the MARBLE benchmark. Keywords: lyrics2song, song generation, long-form, foundation model, music generation

eess.AS↗

arXiv2025

AnyGPT: Unified Multimodal LLM with Discrete Sequence Modeling

We introduce AnyGPT, an any-to-any multimodal language model that utilizes discrete representations for the unified processing of various modalities, including speech, text, images, and music. AnyGPT can be trained stably without any alterations to the current large language model (LLM) architecture or training paradigms. Instead, it relies exclusively on data-level preprocessing, facilitating the seamless integration of new modalities into LLMs, akin to the incorporation of new languages. We build a multimodal text-centric dataset for multimodal alignment pre-training. Utilizing generative models, we synthesize the first large-scale any-to-any multimodal instruction dataset. It consists of 108k samples of multi-turn conversations that intricately interweave various modalities, thus equipping the model to handle arbitrary combinations of multimodal inputs and outputs. Experimental results demonstrate that AnyGPT is capable of facilitating any-to-any multimodal conversation while achieving performance comparable to specialized models across all modalities, proving that discrete representations can effectively and conveniently unify multiple modalities within a language model. Demos are shown in https://junzhan2000.github.io/AnyGPT.github.io/

cs.CL↗

arXiv2025

Impact of Nonlocal Coulomb Repulsion on Superconductivity and Density-Wave Orders in Bilayer Nickelates

The recent discovery of high-temperature superconductivity in pressurized bilayer nickelate La$_3$Ni$_2$O$_7$ and its thin films has generated significant interest in uncovering the underlying pairing mechanisms and correlated electronic states. While earlier theoretical studies have mainly focused on onsite Coulomb interactions, the role of nonlocal Coulomb repulsion has remained largely unexplored. In this work, we systematically investigate the effects of nonlocal Coulomb interactions, in the presence of onsite interactions, on both superconducting and density-wave instabilities using the functional renormalization group (FRG) approach. We find that the interlayer intraorbital repulsion suppresses the interlayer intraorbital $s_{\pm}$-wave pairing and spin-density-wave (SDW) order, while promoting a transition to an interlayer interorbital $d_{x^2-y^2}$-wave pairing state and a mirror-symmetry-breaking charge order. Remarkably, the critical scale of the interorbital $d_{x^2-y^2}$-wave superconductivity is significantly lower than that of the intraorbital $s_{\pm}$-wave superconductivity, indicating that the former is unlikely to account for the observed high-$T_c$ superconductivity. Moreover, the interlayer interorbital repulsion suppresses this $d_{x^2-y^2}$-wave pairing but enhances the $s_{\pm}$-wave pairing through strengthened interlayer charge fluctuations. In addition, the intralayer nearest-neighbor repulsion favors an in-plane charge-density-wave (CDW) order with wave vector $(π,π)$. Our findings reveal the profound impact of nonlocal Coulomb repulsion and underscore the robustness of interlayer pairing rooted in the bilayer structure and multi-orbital nature, thereby advancing the understanding of the intricate correlation effects in bilayer nickelates.

cond-mat.supr-con↗

arXiv2025

Data Mixing Laws: Optimizing Data Mixtures by Predicting Language Modeling Performance

Pretraining data of large language models composes multiple domains (e.g., web texts, academic papers, codes), whose mixture proportions crucially impact the competence of outcome models. While existing endeavors rely on heuristics or qualitative strategies to tune the proportions, we discover the quantitative predictability of model performance regarding the mixture proportions in function forms, which we refer to as the data mixing laws. Fitting such functions on sample mixtures unveils model performance on unseen mixtures before actual runs, thus guiding the selection of an ideal data mixture. Furthermore, we propose nested use of the scaling laws of training steps, model sizes, and our data mixing law to enable predicting the performance of large models trained on massive data under various mixtures with only small-scale training. Moreover, experimental results verify that our method effectively optimizes the training mixture of a 1B model trained for 100B tokens in RedPajama, reaching a performance comparable to the one trained for 48% more steps on the default mixture. Extending the application of data mixing laws to continual training accurately predicts the critical mixture proportion that avoids catastrophic forgetting and outlooks the potential for dynamic data schedules

cs.CL↗

arXiv2024

Cooperation between electron-phonon coupling and electronic interaction in bilayer nickelates La$_3$Ni$_2$O$_7$

The recent observation of high-T$_c$ superconductivity in the bilayer nickelate La$_3$Ni$_2$O$_7$ under pressure has garnered significant interests. While researches have predominantly focused on the role of electron-electron interactions in the superconducting mechanism, the impact of electron-phonon coupling (EPC) has remained elusive. In this work, we perform first-principles calculations to study the phonon spectrum and electron-phonon coupling within La$_3$Ni$_2$O$_7$ under pressure and explore of the interplay between EPC and electronic interactions on the superconductivity by employing functional renormalization group approach. Our calculations reveal that EPC alone is insufficient to trigger superconductivity in La$_3$Ni$_2$O$_7$ under pressure. We identify unique out-of-plane and in-plane breathing phonon modes which selectively couple with the Ni $d_{z^2}$ and $d_{x^2-y^2}$ orbitals, showcasing an orbital-selective EPC. Within the bilayer two-orbital model, it is revealed that solely electronic interactions foster $s_{\pm}$-wave pairing characterized by notable frustration in the band space, leading to a low transition temperature. Remarkably, we find that this out-of-plane EPC can act in concert with electronic interactions to promote the onsite and interlayer pairing in the $d_{z^2}$ orbital, partially releasing the pairing frustration and thus elevating T$_c$. In contrast, the inclusion of in-plane EPC only marginally affects the superconductivity, distinct from the cuprates. Potential experimental implications in La$_3$Ni$_2$O$_7$ are also discussed.

cond-mat.supr-con↗

arXiv2024

SpeechGPT-Gen: Scaling Chain-of-Information Speech Generation

Benefiting from effective speech modeling, current Speech Large Language Models (SLLMs) have demonstrated exceptional capabilities in in-context speech generation and efficient generalization to unseen speakers. However, the prevailing information modeling process is encumbered by certain redundancies, leading to inefficiencies in speech generation. We propose Chain-of-Information Generation (CoIG), a method for decoupling semantic and perceptual information in large-scale speech generation. Building on this, we develop SpeechGPT-Gen, an 8-billion-parameter SLLM efficient in semantic and perceptual information modeling. It comprises an autoregressive model based on LLM for semantic information modeling and a non-autoregressive model employing flow matching for perceptual information modeling. Additionally, we introduce the novel approach of infusing semantic information into the prior distribution to enhance the efficiency of flow matching. Extensive experimental results demonstrate that SpeechGPT-Gen markedly excels in zero-shot text-to-speech, zero-shot voice conversion, and speech-to-speech dialogue, underscoring CoIG's remarkable proficiency in capturing and modeling speech's semantic and perceptual dimensions. Code and models are available at https://github.com/0nutation/SpeechGPT.

cs.CL↗

arXiv2023

SpeechGPT: Empowering Large Language Models with Intrinsic Cross-Modal Conversational Abilities

Multi-modal large language models are regarded as a crucial step towards Artificial General Intelligence (AGI) and have garnered significant interest with the emergence of ChatGPT. However, current speech-language models typically adopt the cascade paradigm, preventing inter-modal knowledge transfer. In this paper, we propose SpeechGPT, a large language model with intrinsic cross-modal conversational abilities, capable of perceiving and generating multi-model content. With discrete speech representations, we first construct SpeechInstruct, a large-scale cross-modal speech instruction dataset. Additionally, we employ a three-stage training strategy that includes modality-adaptation pre-training, cross-modal instruction fine-tuning, and chain-of-modality instruction fine-tuning. The experimental results demonstrate that SpeechGPT has an impressive capacity to follow multi-modal human instructions and highlight the potential of handling multiple modalities with one model. Demos are shown in https://0nutation.github.io/SpeechGPT.github.io/.

cs.CL↗

arXiv2022

HFN: Heterogeneous Feature Network for Multivariate Time Series Anomaly Detection

Network or physical attacks on industrial equipment or computer systems may cause massive losses. Therefore, a quick and accurate anomaly detection (AD) based on monitoring data, especially the multivariate time-series (MTS) data, is of great significance. As the key step of anomaly detection for MTS data, learning the relations among different variables has been explored by many approaches. However, most of the existing approaches do not consider the heterogeneity between variables, that is, different types of variables (continuous numerical variables, discrete categorical variables or hybrid variables) may have different and distinctive edge distributions. In this paper, we propose a novel semi-supervised anomaly detection framework based on a heterogeneous feature network (HFN) for MTS, learning heterogeneous structure information from a mass of unlabeled time-series data to improve the accuracy of anomaly detection, and using attention coefficient to provide an explanation for the detected anomalies. Specifically, we first combine the embedding similarity subgraph generated by sensor embedding and feature value similarity subgraph generated by sensor values to construct a time-series heterogeneous graph, which fully utilizes the rich heterogeneous mutual information among variables. Then, a prediction model containing nodes and channel attentions is jointly optimized to obtain better time-series representations. This approach fuses the state-of-the-art technologies of heterogeneous graph structure learning (HGSL) and representation learning. The experiments on four sensor datasets from real-world applications demonstrate that our approach detects the anomalies more accurately than those baseline approaches, thus providing a basis for the rapid positioning of anomalies.

cs.LG↗

arXiv2022

An Unsupervised Short- and Long-Term Mask Representation for Multivariate Time Series Anomaly Detection

Anomaly detection of multivariate time series is meaningful for system behavior monitoring. This paper proposes an anomaly detection method based on unsupervised Short- and Long-term Mask Representation learning (SLMR). The main idea is to extract short-term local dependency patterns and long-term global trend patterns of the multivariate time series by using multi-scale residual dilated convolution and Gated Recurrent Unit(GRU) respectively. Furthermore, our approach can comprehend temporal contexts and feature correlations by combining spatial-temporal masked self-supervised representation learning and sequence split. It considers the importance of features is different, and we introduce the attention mechanism to adjust the contribution of each feature. Finally, a forecasting-based model and a reconstruction-based model are integrated to focus on single timestamp prediction and latent representation of time series. Experiments show that the performance of our method outperforms other state-of-the-art models on three real-world datasets. Further analysis shows that our method is good at interpretability.

cs.LG↗

arXiv2020

Aerodynamic Data Predictions Based on Multi-task Learning

The quality of datasets is one of the key factors that affect the accuracy of aerodynamic data models. For example, in the uniformly sampled Burgers' dataset, the insufficient high-speed data is overwhelmed by massive low-speed data. Predicting high-speed data is more difficult than predicting low-speed data, owing to that the number of high-speed data is limited, i.e. the quality of the Burgers' dataset is not satisfactory. To improve the quality of datasets, traditional methods usually employ the data resampling technology to produce enough data for the insufficient parts in the original datasets before modeling, which increases computational costs. Recently, the mixtures of experts have been used in natural language processing to deal with different parts of sentences, which provides a solution for eliminating the need for data resampling in aerodynamic data modeling. Motivated by this, we propose the multi-task learning (MTL), a datasets quality-adaptive learning scheme, which combines task allocation and aerodynamic characteristics learning together to disperse the pressure of the entire learning task. The task allocation divides a whole learning task into several independent subtasks, while the aerodynamic characteristics learning learns these subtasks simultaneously to achieve better precision. Two experiments with poor quality datasets are conducted to verify the data quality-adaptivity of the MTL to datasets. The results show than the MTL is more accurate than FCNs and GANs in poor quality datasets.

cs.LG↗

arXiv2015

Deep Speech 2: End-to-End Speech Recognition in English and Mandarin

We show that an end-to-end deep learning approach can be used to recognize either English or Mandarin Chinese speech--two vastly different languages. Because it replaces entire pipelines of hand-engineered components with neural networks, end-to-end learning allows us to handle a diverse variety of speech including noisy environments, accents and different languages. Key to our approach is our application of HPC techniques, resulting in a 7x speedup over our previous system. Because of this efficiency, experiments that previously took weeks now run in days. This enables us to iterate more quickly to identify superior architectures and algorithms. As a result, in several cases, our system is competitive with the transcription of human workers when benchmarked on standard datasets. Finally, using a technique called Batch Dispatch with GPUs in the data center, we show that our system can be inexpensively deployed in an online setting, delivering low latency when serving users at scale.

cs.CL↗