arXiv subjects

Yulong Liu

Publications and source records attributed to Yulong Liu.

At least 37 records · Page 2Linked to original sources

arXiv2025

Foundation Cures Personalization: Improving Personalized Models' Prompt Consistency via Hidden Foundation Knowledge

Facial personalization faces challenges to maintain identity fidelity without disrupting the foundation model's prompt consistency. The mainstream personalization models employ identity embedding to integrate identity information within the attention mechanisms. However, our preliminary findings reveal that identity embeddings compromise the effectiveness of other tokens in the prompt, thereby limiting high prompt consistency and attribute-level controllability. Moreover, by deactivating identity embedding, personalization models still demonstrate the underlying foundation models' ability to control facial attributes precisely. It suggests that such foundation models' knowledge can be leveraged to cure the ill-aligned prompt consistency of personalization models. Building upon these insights, we propose FreeCure, a framework that improves the prompt consistency of personalization models with their latent foundation models' knowledge. First, by setting a dual inference paradigm with/without identity embedding, we identify attributes (e.g., hair, accessories, etc.) for enhancements. Second, we introduce a novel foundation-aware self-attention module, coupled with an inversion-based process to bring well-aligned attribute information to the personalization process. Our approach is training-free, and can effectively enhance a wide array of facial attributes; and it can be seamlessly integrated into existing popular personalization models based on both Stable Diffusion and FLUX. FreeCure has consistently shown significant improvements in prompt consistency across these facial personalization models while maintaining the integrity of their original identity fidelity.

cs.CV↗

arXiv2025

Strong coupling between a single-photon and a two-photon Fock state

The realization of strong nonlinear coupling between single photons has been a long-standing goal in quantum optics and quantum information science, promising wide impact applications, such as all-optical deterministic quantum logic and single-photon frequency conversion. Here, we report an experimental observation of the strong coupling between a single-photon and a two-photon Fock state in an ultrastrongly-coupled circuit-QED system. This strong nonlinear interaction is realized by introducing a detuned flux qubit working as an effective coupler between two modes of a superconducting coplanar waveguide resonator. The ultrastrong light--matter interaction breaks the excitation number conservation, and an external flux bias breaks the parity conservation. The combined effect of the two enables the strong one--two-photon coupling. Quantum Rabi-like avoided crossing is resolved when tuning the two-photon resonance frequency of the first mode across the single-photon resonance frequency of the second mode. Within this new photonic regime, we observe the thresholdless second harmonic generation for a mean photon number below one. Our results represent a key step towards a new regime of quantum nonlinear optics, where individual photons can deterministically and coherently interact with each other in the absence of any stimulating fields.

quant-ph↗

arXiv2025

Optimal Form Factors for Experimental Proposals on Gravity-Induced Entanglement

The interface between quantum mechanics and gravity remains an unresolved issue. Recent advances in precision measurement suggest that detecting gravity-induced entanglement in oscillator systems could provide key evidence for the quantum nature of gravity. However, thermal decoherence imposes strict constraints on system parameters. For entanglement to occur, mechanical frequency $ω_m$, dissipation rate $γ_m$, environmental temperature $T$, oscillator density $ρ$, and the form factor $Λ$-determined by the geometry and arrangement of oscillators-must satisfy a specific constraint. This constraint, intrinsic to the noise model, is considered universal and cannot be improved by quantum control. Given the difficulty in further optimizing $ω_m$, $γ_m$, $ρ$, and $T$, optimizing $Λ$ can relax the constraints on these parameters. In this work, we prove that the form factor has a supremum of $2π$, revealing a fundamental limit of the oscillator system. We propose designs that approach this supremum, nearly an order of magnitude higher than typical spherical oscillators. This optimization could ease experimental constraints and bring quantum gravity validation based on gravity-induced entanglement closer to realization.

quant-ph↗

arXiv2025

Quantize More, Lose Less: Autoregressive Generation from Residually Quantized Speech Representations

Text-to-speech (TTS) synthesis has seen renewed progress under the discrete modeling paradigm. Existing autoregressive approaches often rely on single-codebook representations, which suffer from significant information loss. Even with post-hoc refinement techniques such as flow matching, these methods fail to recover fine-grained details (e.g., prosodic nuances, speaker-specific timbres), especially in challenging scenarios like singing voice or music synthesis. We propose QTTS, a novel TTS framework built upon our new audio codec, QDAC. The core innovation of QDAC lies in its end-to-end training of an ASR-based auto-regressive network with a GAN, which achieves superior semantic feature disentanglement for scalable, near-lossless compression. QTTS models these discrete codes using two innovative strategies: the Hierarchical Parallel architecture, which uses a dual-AR structure to model inter-codebook dependencies for higher-quality synthesis, and the Delay Multihead approach, which employs parallelized prediction with a fixed delay to accelerate inference speed. Our experiments demonstrate that the proposed framework achieves higher synthesis quality and better preserves expressive content compared to baseline. This suggests that scaling up compression via multi-codebook modeling is a promising direction for high-fidelity, general-purpose speech and audio generation.

cs.SD↗

arXiv2025

Experimental and numerical study on current distribution in parallel co-wound no-insulation coils

No-insulation (NI) coils are known for their high thermal stability and self-protection features due to turn-to-turn contacts. Parallel co-winding is a promising method to reduce the charging delay of NI coils while maintaining thermal stability, demonstrating significant potential for applications in fusion and other large-scale or high-field magnets. The non-uniform current distribution among parallel superconducting tapes in parallel co-wound NI coils may lead to thermal and mechanical stability issues. In this work, we conducted current measurement experiments on small parallel co-wound NI REBCO coils to investigate the non-uniform current distribution and its influencing factors. The parallel tapes in the input and output sections of the test coils were separated and a series of Rogowski coils was used to measure the current in each tape during ramping charging process. We combined a field-circuit coupled model based on the T-A formulation with an equivalent circuit model to calculate the current distribution in co-wound coils. Both the measured and calculated results indicated that the current distribution during ramping was highly non-uniform, with some tapes carrying reverse currents. We calculated the current distribution in co-wound coils with different insulation methods and analyzed the influencing factors of the reverse current. The influence of the terminal resistance on current distribution was also discussed. This work could contribute to a deeper understanding of current distribution behavior in co-wound coils and provide insights for their application in large-scale or high-field magnet systems.

physics.app-ph↗

arXiv2025

Superior Frequency Stability and Long-Lived State-Swapping in Cubic-SiC Mechanical Mode Pairs

The multimode cavity optomechanical system offers versatile applications including state transduction, coherent interconnection, and many-body simulations. In this study, we developed a cavity electromechanical system that integrates a 3C-SiC membrane and a rectangular superconducting cavity to observe the generation of nearly degenerate pairs of mechanical modes. Subsequently, we derive the expression for intrinsic frequency under nonuniform stress and find that this method supports a remarkably resolution for stress analysis in thin films. Experimentally, we perform collective fitting on the measured set of 57 mechanical modes, revealing deviations in biaxial non-uniform stress on the order of MPa. These degeneracy-broken mechanical modes exhibit exceptional quality factors as high as $10^8$ in a thermal bath of 10 mK. Furthermore, Allan deviation indicates that these modes exhibit extremely stable frequencies compared with different types of optomechanical devices. We then performed state-swapping between near-degenerate mode pairs, demonstrating the transfer efficiency exceeding 78\%, attributed to their exceptionally long lifetimes. This study paves the way for the design of compact quantum phononic devices featuring high-quality-factor mechanical multimode.

quant-ph↗

arXiv2025

Amplitude Noise Cancellation of Microwave Tones

Carrier noise in coherent tones limits sensitivity and causes heating in many experimental systems, such as force sensors, time-keeping, and studies of macroscopic quantum phenomena. Much progress has been made to reduce carrier noise using phase noise cancellation techniques, however, in systems where amplitude noise dominates, these methods are ineffective. Here, we present a technique to reduce amplitude noise from microwave generators using feedback cancellation. The method uses a field-programmable gate array (FPGA) to reproduce noise with a tunable gain and time delay, resulting in destructive interference when combined with the original tone. The FPGA additionally allows for tuning of the frequency offset and bandwidth in which the noise is canceled. By employing the cancellation we observe 13 dB of noise power reduction at a 2 MHz offset from a 4 GHz microwave tone, lowering the total noise to the phase noise level. To verify its applicability we utilize the setup in a microwave optomechanics experiment to investigate the effect of generator noise on the sideband cooling of a 0.5 mm silicon nitride membrane resonator. We observe that with our technique the rate of externally induced cavity heating is reduced by a factor of 3.5 and the minimum oscillator occupation is lowered by a factor of 2. This method broadens the field of noise cancellation techniques, where amplitude noise is becoming an increasingly important consideration in microwave systems as phase noise performances improve over time.

physics.app-ph↗

arXiv2025

Degeneracy-breaking and Long-lived Multimode Microwave Electromechanical Systems Enabled by Cubic Silicon-Carbide Membrane Crystals

Cubic silicon-carbide crystals (3C-SiC), known for their high thermal conductivity and in-plane stress, hold significant promise for the development of high-quality ($Q$) mechanical oscillators. We reveal degeneracy-breaking phenomena in 3C-phase crystalline silicon-carbide membrane and present high-$Q$ mechanical modes in pairs or clusters. The 3C-SiC material demonstrates excellent microwave compatibility with superconducting circuits. Thus, we can establish a coherent electromechanical interface, enabling precise control over 21 high-$Q$ mechanical modes from a single 3C-SiC square membrane. Benefiting from extremely high mechanical frequency stability, this interface enables tunable light slowing with group delays extending up to an impressive duration of \emph{an hour}. Coherent energy transfer between distinct mechanical modes are also presented. In this work, the studied 3C-SiC membrane crystal with their significant properties of multiple acoustic modes and high-quality factors, provide unique opportunities for the encoding, storage, and transmission of quantum information via bosonic phonon channels.

quant-ph↗

arXiv2024

Adiabatic sheath model for beam-driven blowout plasma channels

In plasma wakefield accelerators, the structure of the blowout sheath is vital for the blowout radius and the electromagnetic field distribution inside the blowout. Previous theories assume artificial distribution functions for the sheath, which are either inaccurate or require prior knowledge of parameters. In this study, we develop an adiabatic sheath model based on force balancing, which leads to a self-consistent form of the sheath distribution. This model gives a better estimate of the blowout channel balancing radius than previous models.

physics.plasm-ph↗

arXiv2024

See Through Their Minds: Learning Transferable Neural Representation from Cross-Subject fMRI

Deciphering visual content from functional Magnetic Resonance Imaging (fMRI) helps illuminate the human vision system. However, the scarcity of fMRI data and noise hamper brain decoding model performance. Previous approaches primarily employ subject-specific models, sensitive to training sample size. In this paper, we explore a straightforward but overlooked solution to address data scarcity. We propose shallow subject-specific adapters to map cross-subject fMRI data into unified representations. Subsequently, a shared deeper decoding model decodes cross-subject features into the target feature space. During training, we leverage both visual and textual supervision for multi-modal brain decoding. Our model integrates a high-level perception decoding pipeline and a pixel-wise reconstruction pipeline guided by high-level perceptions, simulating bottom-up and top-down processes in neuroscience. Empirical experiments demonstrate robust neural representation learning across subjects for both pipelines. Moreover, merging high-level and low-level information improves both low-level and high-level reconstruction metrics. Additionally, we successfully transfer learned general knowledge to new subjects by training new adapters with limited training data. Compared to previous state-of-the-art methods, notably pre-training-based methods (Mind-Vis and fMRI-PTE), our approach achieves comparable or superior results across diverse tasks, showing promise as an alternative method for cross-subject fMRI data pre-training. Our code and pre-trained weights will be publicly released at https://github.com/YulongBonjour/See_Through_Their_Minds.

cs.CV↗

arXiv2024

Gate-Compatible Circuit Quantum Electrodynamics in a Three-Dimensional Cavity Architecture

Semiconductor-based superconducting qubits offer a versatile platform for studying hybrid quantum devices in circuit quantum electrodynamics (cQED) architecture. Most of these cQED experiments utilize coplanar waveguides, where the incorporation of DC gate lines is straightforward. Here, we present a technique for probing gate-tunable hybrid devices using a three-dimensional (3D) microwave cavity. A recess is machined inside the cavity wall for the placement of devices and gate lines. We validate this design using a hybrid device based on an InAs-Al nanowire Josephson junction. The coupling between the device and the cavity is facilitated by a long superconducting strip, the antenna. The Josephson junction and the antenna together form a gatemon qubit. We further demonstrate the gate-tunable cavity shift and two-tone qubit spectroscopy. This technique could be used to probe various quantum devices and materials in a 3D cQED architecture that requires DC gate voltages.

quant-ph↗

arXiv2024

Near-ground state cooling in electromechanics using measurement-based feedback and Josephson parametric amplifier

Feedback-based control of nano- and micromechanical resonators can enable the study of macroscopic quantum phenomena and also sensitive force measurements. Here, we demonstrate the feedback cooling of a low-loss and high-stress macroscopic SiN membrane resonator close to its quantum ground state. We use the microwave optomechanical platform, where the resonator is coupled to a microwave cavity. The experiment utilizes a Josephson travelling wave parametric amplifier, which is nearly quantum-limited in added noise, and is important to mitigate resonator heating due to system noise in the feedback loop. We reach a thermal phonon number as low as 1.6, which is limited primarily by microwave-induced heating. We also discuss the sideband asymmetry observed when a weak microwave tone for independent readout is applied in addition to other tones used for the cooling. The asymmetry can be qualitatively attributed to the quantum-mechanical imbalance between emission and absorption. However, we find that the observed asymmetry is only partially due to this quantum effect. In specific situations, the asymmetry is fully dominated by a cavity Kerr effect under multitone irradiation.

quant-ph↗

arXiv2024

From Summary to Action: Enhancing Large Language Models for Complex Tasks with Open World APIs

The distinction between humans and animals lies in the unique ability of humans to use and create tools. Tools empower humans to overcome physiological limitations, fostering the creation of magnificent civilizations. Similarly, enabling foundational models like Large Language Models (LLMs) with the capacity to learn external tool usage may serve as a pivotal step toward realizing artificial general intelligence. Previous studies in this field have predominantly pursued two distinct approaches to augment the tool invocation capabilities of LLMs. The first approach emphasizes the construction of relevant datasets for model fine-tuning. The second approach, in contrast, aims to fully exploit the inherent reasoning abilities of LLMs through in-context learning strategies. In this work, we introduce a novel tool invocation pipeline designed to control massive real-world APIs. This pipeline mirrors the human task-solving process, addressing complicated real-life user queries. At each step, we guide LLMs to summarize the achieved results and determine the next course of action. We term this pipeline `from Summary to action', Sum2Act for short. Empirical evaluations of our Sum2Act pipeline on the ToolBench benchmark show significant performance improvements, outperforming established methods like ReAct and DFSDT. This highlights Sum2Act's effectiveness in enhancing LLMs for complex real-world tasks.

cs.AI↗

arXiv2023

Uneven illumination surface defects inspection based on convolutional neural network

Surface defect inspection based on machine vision is often affected by uneven illumination. In order to improve the inspection rate of surface defects inspection under uneven illumination condition, this paper proposes a method for detecting surface image defects based on convolutional neural network, which is based on the adjustment of convolutional neural networks, training parameters, changing the structure of the network, to achieve the purpose of accurately identifying various defects. Experimental on defect inspection of copper strip and steel images shows that the convolutional neural network can automatically learn features without preprocessing the image, and correct identification of various types of image defects affected by uneven illumination, thus overcoming the drawbacks of traditional machine vision inspection methods under uneven illumination.

cs.CV↗

arXiv2023

BrainCLIP: Bridging Brain and Visual-Linguistic Representation Via CLIP for Generic Natural Visual Stimulus Decoding

Due to the lack of paired samples and the low signal-to-noise ratio of functional MRI (fMRI) signals, reconstructing perceived natural images or decoding their semantic contents from fMRI data are challenging tasks. In this work, we propose, for the first time, a task-agnostic fMRI-based brain decoding model, BrainCLIP, which leverages CLIP's cross-modal generalization ability to bridge the modality gap between brain activity, image, and text. Our experiments demonstrate that CLIP can act as a pivot for generic brain decoding tasks, including zero-shot visual categories decoding, fMRI-image/text matching, and fMRI-to-image generation. Specifically, BrainCLIP aims to train a mapping network that transforms fMRI patterns into a well-aligned CLIP embedding space by combining visual and textual supervision. Our experiments show that this combination can boost the decoding model's performance on certain tasks like fMRI-text matching and fMRI-to-image generation. On the zero-shot visual category decoding task, BrainCLIP achieves significantly better performance than BraVL, a recently proposed multi-modal method specifically designed for this task. BrainCLIP can also reconstruct visual stimuli with high semantic fidelity and establishes a new state-of-the-art for fMRI-based natural image reconstruction in terms of high-level semantic features.

cs.CV↗

arXiv2023

Image Super-Resolution using Efficient Striped Window Transformer

Transformers have achieved remarkable results in single-image super-resolution (SR). However, the challenge of balancing model performance and complexity has hindered their application in lightweight SR (LSR). To tackle this challenge, we propose an efficient striped window transformer (ESWT). We revisit the normalization layer in the transformer and design a concise and efficient transformer structure to build the ESWT. Furthermore, we introduce a striped window mechanism to model long-term dependencies more efficiently. To fully exploit the potential of the ESWT, we propose a novel flexible window training strategy that can improve the performance of the ESWT without additional cost. Extensive experiments show that ESWT outperforms state-of-the-art LSR transformers, and achieves a better trade-off between model performance and complexity. The ESWT requires fewer parameters, incurs faster inference, smaller FLOPs, and less memory consumption, making it a promising solution for LSR.

cs.CV↗

arXiv2023

Gatemon qubit based on a thin InAs-Al hybrid nanowire

We study a gate-tunable superconducting qubit (gatemon) based on a thin InAs-Al hybrid nanowire. Using a gate voltage to control its Josephson energy, the gatemon can reach the strong coupling regime to a microwave cavity. In the dispersive regime, we extract the energy relaxation time $T_1\sim$0.56 $μ$s and the dephasing time $T_2^* \sim$0.38 $μ$s. Since thin InAs-Al nanowires can have fewer or single sub-band occupation and recent transport experiment shows the existence of nearly quantized zero-bias conductance peaks, our result holds relevancy for detecting Majorana zero modes in thin InAs-Al nanowires using circuit quantum electrodynamics.

cond-mat.mes-hall↗

arXiv2022

Parity-dependent unidirectional and chiral photon transfer in reversed-dissipation cavity optomechanics

Nonreciprocal elements, such as isolators and circulators, play an important role in classical and quantum information processing. Recently, strong nonreciprocal effects have been experimentally demonstrated in cavity optomechanical systems. In these approaches, the bandwidth of the nonreciprocal photon transmission is limited by the mechanical resonator linewidth, which is arguably much smaller than the linewidths of the cavity modes in most electromechanical or optomechanical devices. In this work, we demonstrate broadband nonreciprocal photon transmission in the \emph{reversed-dissipation} regime, where the mechanical mode with a large decay rate can be adiabatically eliminated while mediating anti-$\mathcal{PT}$-symmetric dissipative coupling with two kinds of phase factors. Adjusting the relative phases allows the observation of \emph{periodic} Riemann-sheet structures with distributed exceptional points (Eps). At the Eps, destructive quantum interference breaks both the $\mathcal{T}$- and $\mathcal{P}$-inversion symmetry, resulting in unidirectional and chiral photon transmissions. In the reversed-dissipation regime, the nonreciprocal bandwidth is no longer limited by the mechanical mode linewidth but is improved to the linewidth of the cavity resonance. Furthermore, we find that the direction of the unidirectional and chiral energy transfer could be reversed by changing the \emph{parity} of the Eps. Extending non-Hermitian couplings to a three-cavity model, the broken anti-$\mathcal{PT}$-symmetry allows us to observe high-order Eps, at which a parity-dependent chiral circulator is demonstrated. The driving-phase controlled periodical Riemann sheets allow observation of the parity-dependent unidirectional and chiral energy transfer and thus provide a useful cell for building up nonreciprocal array and realizing topological, e.g., isolators, circulators, or amplifiers.

quant-ph↗