Search

Several upper-limb exoskeleton robots have been developed for stroke rehabilitation, but their rather low level of individualized assistance typically limits their effectiveness and practicability. Individualized assistance involves an upper-limb exoskeleton robot continuously assessing feedback from a stroke patient and then meticulously adjusting interaction forces to suit specific conditions and online changes. This paper describes the development of a new upper-limb exoskeleton robot with a novel online generative capability that allows it to provide individualized assistance to support the rehabilitation training of stroke patients. Specifically, the upper-limb exoskeleton robot exploits generative models to customize the fine and fit trajectory for the patient, as medical conditions, responses, and comfort feedback during training generally differ between patients. This generative capability is integrated into the two working modes of the upper-limb exoskeleton robot: an active mirroring mode for patients who retain motor abilities on one side of the body and a passive following mode for patients who lack motor ability on both sides of the body. The performance of the upper-limb exoskeleton robot was illustrated in experiments involving healthy subjects and stroke patients.

Quasi-periodic oscillations in solar flaring emission have been observed over the past few decades. To date, the underpinning processes resulting in the quasi-periodic oscillations remain unknown. In this paper, we report a unique event that exhibits both the long-duration quasi-periodic intensity oscillations of flare loops and the quasi-periodic slipping motion of ribbon substructures during a C9.1-class flare (SOL2015-03-15-T01:15), using the observations from Solar Dynamics Observatory and Interface Region Imaging Spectrograph. The high-temperature flare loops rooted in the straight part of ribbons display a "bright-dim" intensity oscillation, with a period of about 4.5 minutes. The oscillation starts just after the flare onset and lasts over 3 hours. Meanwhile, the substructures within the ribbon tip display the quasi-periodic slipping motion along the ribbon at 1400 \AA images which has a similar periodicity to the stationary intensity oscillation of the flare loops in the straight part of the flare ribbons. We suggest that the quasi-periodic pattern is probably related to the loop-top dynamics caused by the reconnection outflow impinging on the flare loops., Comment: 13 pages, 10 figures, accepted for publication in ApJ

Large Language Models (LLMs) have played an important role in many fields due to their powerful capabilities.However, their massive number of parameters leads to high deployment requirements and incurs significant inference costs, which impedes their practical applications. Training smaller models is an effective way to address this problem. Therefore, we introduce OpenBA-V2, a 3.4B model derived from multi-stage compression and continual pre-training from the original 15B OpenBA model. OpenBA-V2 utilizes more data, more flexible training objectives, and techniques such as layer pruning, neural pruning, and vocabulary pruning to achieve a compression rate of 77.3\% with minimal performance loss. OpenBA-V2 demonstrates competitive performance compared to other open-source models of similar size, achieving results close to or on par with the 15B OpenBA model in downstream tasks such as common sense reasoning and Named Entity Recognition (NER). OpenBA-V2 illustrates that LLMs can be compressed into smaller ones with minimal performance loss by employing advanced training objectives and data strategies, which may help deploy LLMs in resource-limited scenarios.

Graphene based nanomaterials have attracted significant attention for their potentials in biomedical and biotechnology applications in recent years, owing to the outstanding physical and chemical properties. However, the interaction mechanism and impact on biological activity of macro and micro biomolecules still require more concerns and further research in order to enhance their applicability in biosensors, etc. Herein, an integrated method has been developed to predict the protein bioactivity performance when interacting with nanomaterials for protein based biosensor. Molecular dynamics simulation and molecular docking technique were consolidated to investigate several nanomaterials C60 fullerene, single walled carbon nanotube, pristine graphene and graphene oxide, and their effect when interacting with protein. The adsorption behavior, secondary structure changes and protein bioactivity changes were simulated, and the results of protein activity simulation were verified in combination with atomic force spectrum, circular dichroism spectrum fluorescence and electrochemical experiments. The best quantification alignment between bioactivity obtained by simulation and experiment measurements was further explored. The two proteins, RNase A and Exonuclease III, were regarded as analysis model for the proof of concept, and the prediction accuracy of protein bioactivty could reach up to 0.98.

Decoding continuous language from brain activity is a formidable yet promising field of research. It is particularly significant for aiding people with speech disabilities to communicate through brain signals. This field addresses the complex task of mapping brain signals to text. The previous best attempt reverse-engineered this process in an indirect way: it began by learning to encode brain activity from text and then guided text generation by aligning with predicted brain responses. In contrast, we propose a simple yet effective method that guides text reconstruction by directly comparing them with the predicted text embeddings mapped from brain activities. Comprehensive experiments reveal that our method significantly outperforms the current state-of-the-art model, showing average improvements of 77% and 54% on BLEU and METEOR scores. We further validate the proposed modules through detailed ablation studies and case analyses and highlight a critical correlation: the more precisely we map brain activities to text embeddings, the better the text reconstruction results. Such insight can simplify the task of reconstructing language from brain activities for future work, emphasizing the importance of improving brain-to-text-embedding mapping techniques., Comment: Accepted to NAACL 2024 main conference

Recently, a black hole model in loop quantum gravity has been proposed by Lewandowski, Ma, Yang and Zhang (Phys. Rev. Lett. \textbf{130}, 101501 (2023)). The metric tensor of the quantum black hole (QBH) is a suitably modified Schwarzschild one. In this paper, we calculate the radius of light ring and obtain the linear approximation of it with respect to the quantum correction parameter $\alpha$: $r_{l} \simeq 3 M - \frac{\alpha}{9 M}$. We then assume the QBH is backlit by a large, distant plane of uniform, isotropic emission and calculate the radius of the black hole shadow and its linear approximation: $r_{s} = 3 \sqrt{3} M - \frac{\alpha}{6 \left(\sqrt{3} M\right)}$. We also consider the photon ring structures in the shadow when the impact parameter $b$ of the photon approaches to a critical impact parameter $b_{\textrm{c}}$, and obtain a formula for estimating the deflection angle, which is $\varphi_{\textrm{def}} = - \frac{\sqrt{2}}{\omega r_{l}^2}\log{\left(b - b_c\right) + \widetilde{C}(b)}$. We also numerically plot the images of shadows and photon rings of the QBH in three different illumination models and compare them with that of a Schwarzschild in each model. It is found that we could distinguish the quantum black hole with a Schwarzschild black hole by the shadow images in certain specific illumination model., Comment: references added,one figure added

We propose a mechanism for the excitation of large-scale quasi-periodic fast-propagating magnetoacoustic (QFP) waves observed on both sides of the coronal mass ejection (CME). Through a series of numerical experiments, we successfully simulated the quasi-static evolution of the equilibrium locations of the magnetic flux rope in response to the change of the background magnetic field, as well as the consequent loss of the equilibrium that eventually gives rise to the eruption. During the eruption, we identified QFP waves propagating radially outwards the flux rope, and tracing their origin reveals that they result from the disturbance within the flux rope. Acting as an imperfect waveguide, the flux rope allows the internal disturbance to escape to the outside successively via its surface, invoking the observed QFP waves. Furthermore, we synthesized the images of QFP waves on the basis of the data given by our simulations, and found the consistence with observations. This indicates that the leakage of the disturbance outside the flux rope could be a reasonable mechanism of QFP waves., Comment: 8 figures, accepted by the Astrophysical Journal

The precise spatial and quantitative delineation of indistinct-boundary medical objects is paramount for the accuracy of diagnostic protocols, efficacy of surgical interventions, and reliability of postoperative assessments. Despite their significance, the effective segmentation and instantaneous three-dimensional reconstruction are significantly impeded by the paucity of representative samples in available datasets and noise artifacts. To surmount these challenges, we introduced Stochastic Defect Injection (SDi) to augment the representational diversity of challenging indistinct-boundary objects within training corpora. Consequently, we propose the Dual-Encoder Fourier Group Harmonics Network (DEFN) to tailor noise filtration, amplify detailed feature recognition, and bolster representation across diverse medical imaging scenarios. By incorporating Dynamic Weight Composing (DWC) loss dynamically adjusts model's focus based on training progression, DEFN achieves SOTA performance on the OIMHS public dataset, showcasing effectiveness in indistinct boundary contexts. Source code for DEFN is available at: https://github.com/IMOP-lab/DEFN-pytorch., Comment: 36pages,16figures,7tables

The meaningful correlations between the zero-sound modes and the stiffness of the nuclear equation of state (EOS) are uncovered in nuclear matter with the relativistic mean-field theory. It is demonstrated that the high-density zero-sound modes merely exist in models with the stiff EOS. While the stiff EOS can be softened by including {\omega}-meson self-interactions (the {\omega}4 term), the weakened coupling of the {\omega}-meson self-interactions reignites the zero sound at high density. These results suggest that the high-density zero-sound modes can be used to probe the stiffness of the EOS at supra-normal densities. The implications and effects of zero sounds are also discussed in heavy ion collisions and neutron stars.

One of the typical purposes of using lower-limb exoskeleton robots is to provide assistance to the wearer by supporting their weight and augmenting their physical capabilities according to a given task and human motion intentions. The generalizability of robots across different wearers in multiple tasks is important to ensure that the robot can provide correct and effective assistance in actual implementation. However, most lower-limb exoskeleton robots exhibit only limited generalizability. Therefore, this paper proposes a human-in-the-loop learning and adaptation framework for exoskeleton robots to improve their performance in various tasks and for different wearers. To suit different wearers, an individualized walking trajectory is generated online using dynamic movement primitives and Bayes optimization. To accommodate various tasks, a task translator is constructed using a neural network to generalize a trajectory to more complex scenarios. These generalization techniques are integrated into a unified variable impedance model, which regulates the exoskeleton to provide assistance while ensuring safety. In addition, an anomaly detection network is developed to quantitatively evaluate the wearer's comfort, which is considered in the trajectory learning procedure and contributes to the relaxation of conflicts in impedance control. The proposed framework is easy to implement, because it requires proprioceptive sensors only to perform and deploy data-efficient learning schemes. This makes the exoskeleton practical for deployment in complex scenarios, accommodating different walking patterns, habits, tasks, and conflicts. Experiments and comparative studies on a lower-limb exoskeleton robot are performed to demonstrate the effectiveness of the proposed framework., Comment: 16 pages journal article

While Federated learning (FL) is attractive for pulling privacy-preserving distributed training data, the credibility of participating clients and non-inspectable data pose new security threats, of which poisoning attacks are particularly rampant and hard to defend without compromising privacy, performance or other desirable properties of FL. To tackle this problem, we propose a self-purified FL (SPFL) method that enables benign clients to exploit trusted historical features of locally purified model to supervise the training of aggregated model in each iteration. The purification is performed by an attention-guided self-knowledge distillation where the teacher and student models are optimized locally for task loss, distillation loss and attention-based loss simultaneously. SPFL imposes no restriction on the communication protocol and aggregator at the server. It can work in tandem with any existing secure aggregation algorithms and protocols for augmented security and privacy guarantee. We experimentally demonstrate that SPFL outperforms state-of-the-art FL defenses against various poisoning attacks. The attack success rate of SPFL trained model is at most 3$\%$ above that of a clean model, even if the poisoning attack is launched in every iteration with all but one malicious clients in the system. Meantime, it improves the model quality on normal inputs compared to FedAvg, either under attack or in the absence of an attack.

Self-attention and position embedding are two key modules in transformer-based Large Language Models (LLMs). However, the potential relationship between them is far from well studied, especially for long context window extending. In fact, anomalous behaviors harming long context extrapolation exist between Rotary Position Embedding (RoPE) and vanilla self-attention unveiled by our work. To address this issue, we propose a novel attention mechanism, CoCA (Collinear Constrained Attention). Specifically, we enforce a collinear constraint between $Q$ and $K$ to seamlessly integrate RoPE and self-attention. While only adding minimal computational and spatial complexity, this integration significantly enhances long context window extrapolation ability. We provide an optimized implementation, making it a drop-in replacement for any existing transformer-based models. Extensive experiments show that CoCA performs extraordinarily well in extending context windows. A CoCA-based GPT model, trained with a context length of 512, can seamlessly extend the context window up to 32K (60$\times$), without any fine-tuning. Additionally, by dropping CoCA in LLaMA-7B, we achieve extrapolation up to 32K within only 2K training length. Our code is publicly available at: https://github.com/codefuse-ai/Collinear-Constrained-Attention, Comment: 16 pages, 7 figures

A typical application of upper-limb exoskeleton robots is deployment in rehabilitation training, helping patients to regain manipulative abilities. However, as the patient is not always capable of following the robot, safety issues may arise during the training. Due to the bias in different patients, an individualized scheme is also important to ensure that the robot suits the specific conditions (e.g., movement habits) of a patient, hence guaranteeing effectiveness. To fulfill this requirement, this paper proposes a new motion planning scheme for upper-limb exoskeleton robots, which drives the robot to provide customized, safe, and individualized assistance using both human demonstration and interactive learning. Specifically, the robot first learns from a group of healthy subjects to generate a reference motion trajectory via probabilistic movement primitives (ProMP). It then learns from the patient during the training process to further shape the trajectory inside a moving safe region. The interactive data is fed back into the ProMP iteratively to enhance the individualized features for as long as the training process continues. The robot tracks the individualized trajectory under a variable impedance model to realize the assistance. Finally, the experimental results are presented in this paper to validate the proposed control scheme., Comment: Has been submitted to ICRA_2024

A current sheet (CS) is the central structure in the disrupting magnetic configuration during solar eruptions. More than 90\% of the free magnetic energy (the difference between the energy in the non-potential magnetic field and that in the potential one) stored in the coronal magnetic field beforehand is converted into heating and kinetic energy of the plasma, as well as accelerating charged particles, by magnetic reconnection occurring in the CS. However, the detailed physical properties and fine structures of the CS are still unknown since there is no relevant information obtained via in situ detections. The Parker Solar Probe (PSP) may provide us such information should it traverse a CS in the eruption. The perihelion of PSP's final orbit is located at about 10 solar radii from the center of the Sun, so it can observe the CS at a very close distance, or even traverses the CS, which provides us a unique opportunity to look into fine properties and structures of the CS, helping reveal the detailed physics of large-scale reconnection that was impossible before. We evaluate the probability that PSP can traverse a CS, and examine the orbit of a PSP-like spacecraft that has the highest probability to traverse a CS., Comment: 28 pages, 12 figure, accepted by ApJS

Turbulence plays a key role for forming the complex geometry of the large-scale current sheet (CS) and fast energy release in a solar eruption. In this paper, we present full 3D high-resolution simulations for the process of a moderate Coronal Mass Ejection (CME) and the thermodynamical evolution of the highly confined CS. Copious elongated blobs are generated due to tearing and plasmoid instabilities giving rise to a higher reconnection rate and undergo the splitting, merging and kinking processes in a more complex way in 3D. A detailed thermodynamical analysis shows that the CS is mainly heated by adiabatic and numerical viscous terms, and thermal conduction is the dominant factor that balances the energy inside the CS. Accordingly, the temperature of the CS reaches to a maximum of about 20 MK and the range of temperatures is relatively narrow. From the face-on view in the synthetic Atmospheric Imaging Assembly 131 $\mathring{A}$, the downflowing structures with similar morphology to supra-arcade downflows are mainly located between the post-flare loops and loop-top, while moving blobs can extend spikes higher above the loop-top. The downward-moving plasmoids can keep the twisted magnetic field configuration until the annihilation at the flare loop-top, indicating that plasmoid reconnection dominates in the lower CS. Meanwhile, the upward-moving ones turn into turbulent structures before arriving at the bottom of the CME, implying that turbulent reconnection dominates in the upper CS. The spatial distributions of the turbulent energy and anisotropy are addressed, which show a significant variation in the spectra with height.

We study the neutron star core-crust transition density $\rho_t$ with the inclusion of the vacuum polarization in the dielectric function in the nonlinear relativistic Hartree approach (RHAn). It is found that the strong correlation between the $\rho_{t}$ and the scalar meson mass $m_{\sigma}$ strikingly overwhelms the uncertainty of the nuclear equation of state in the RHAn models, in contrast to the usual awareness that $\rho_{t}$ is predominantly sensitive to the isovector nuclear potential and symmetry energy. The accurate extraction of $\rho_{t}$ through the future gravitational wave measurements can thus provide a strong constraint on the longstanding uncertainty of $m_{\sigma}$, which is of significance to better infer the vacuum property. As an astrophysical implication, it suggests that the correlation between $\rho_t$ and $m_\sigma$ is very favorable to reconcile the difficulty in reproducing the large crustal moment of inertia for the pulsar glitches with the well constrained symmetry energy., Comment: 11 pages, 4 figures

The acceleration and transport of energetic electrons during solar flares is one of the outstanding topics in solar physics. Recent X-ray and radio imaging and spectroscopy observations have provided diagnostics of the distribution of nonthermal electrons and suggested that, in certain flare events, electrons are primarily accelerated in the loop-top and likely experience trapping and/or scattering effects. By combining the focused particle transport equation with magnetohydrodynamic (MHD) simulations of solar flares, we present a macroscopic particle model that naturally incorporates electron acceleration and transport. Our simulation results indicate that the physical processes such as turbulent pitch-angle scattering can have important impacts on both electron acceleration in the loop-top and transport in the flare loop, and their influences are highly energy dependent. A spatial-dependent turbulent scattering with enhancement in the loop-top can enable both efficient electron acceleration to high energies and transport of abundant electrons to the footpoints. We further generate spatially resolved synthetic hard X-ray (HXR) emission images and spectra, revealing both the loop-top and footpoint HXR sources. Similar to the observations, we show that the footpoint HXR sources are brighter and harder than the loop-top HXR source. We suggest that the macroscopic particle model provides new insights into understanding the connection between the observed loop-top and footpoint nonthermal emission sources by combining the particle model with dynamically evolving MHD simulations of solar flares., Comment: Accepted for publication in ApJL

Secure aggregation is widely used in horizontal Federated Learning (FL), to prevent leakage of training data when model updates from data owners are aggregated. Secure aggregation protocols based on Homomorphic Encryption (HE) have been utilized in industrial cross-silo FL systems, one of the settings involved with privacy-sensitive organizations such as financial or medical, presenting more stringent requirements on privacy security. However, existing HE-based solutions have limitations in efficiency and security guarantees against colluding adversaries without a Trust Third Party. This paper proposes an efficient Doubly Homomorphic Secure Aggregation (DHSA) scheme for cross-silo FL, which utilizes multi-key Homomorphic Encryption (MKHE) and seed homomorphic pseudorandom generator (SHPRG) as cryptographic primitives. The application of MKHE provides strong security guarantees against up to $N-2$ participates colluding with the aggregator, with no TTP required. To mitigate the large computation and communication cost of MKHE, we leverage the homomorphic property of SHPRG to replace the majority of MKHE computation by computationally-friendly mask generation from SHPRG, while preserving the security. Overall, the resulting scheme satisfies the stringent security requirements of typical cross-silo FL scenarios, at the same time providing high computation and communication efficiency for practical usage. We experimentally demonstrate our scheme brings a speedup to 20$\times$ over the state-of-the-art HE-based secure aggregation, and reduces the traffic volume to approximately 1.5$\times$ inflation over the plain learning setting., Comment: 34 pages, 6 figures