Search

Large transformer models have demonstrated remarkable success. Post-training quantization (PTQ), which requires only a small dataset for calibration and avoids end-to-end retraining, is a promising solution for compressing these large models. Regrettably, existing PTQ methods typically exhibit non-trivial performance loss. We find that the performance bottleneck stems from over-consideration of hardware compatibility in the quantization process, compelling them to reluctantly employ simple quantizers, albeit at the expense of accuracy. With the above insights, we propose RepQuant, a novel PTQ framework with quantization-inference decoupling paradigm to address the above issues. RepQuant employs complex quantizers in the quantization process and simplified quantizers in the inference process, and performs mathematically equivalent transformations between the two through quantization scale reparameterization, thus ensuring both accurate quantization and efficient inference. More specifically, we focus on two components with extreme distributions: LayerNorm activations and Softmax activations. Initially, we apply channel-wise quantization and log$\sqrt{2}$ quantization, respectively, which are tailored to their distributions. In particular, for the former, we introduce a learnable per-channel dual clipping scheme, which is designed to efficiently identify outliers in the unbalanced activations with fine granularity. Then, we reparameterize the scales to hardware-friendly layer-wise quantization and log2 quantization for inference. Moreover, quantized weight reconstruction is seamlessly integrated into the above procedure to further push the performance limits. Extensive experiments are performed on different large-scale transformer variants on multiple tasks, including vision, language, and multi-modal transformers, and RepQuant encouragingly demonstrates significant performance advantages.

Diffusion models have achieved great success in image generation tasks through iterative noise estimation. However, the heavy denoising process and complex neural networks hinder their low-latency applications in real-world scenarios. Quantization can effectively reduce model complexity, and post-training quantization (PTQ), which does not require fine-tuning, is highly promising in accelerating the denoising process. Unfortunately, we find that due to the highly dynamic distribution of activations in different denoising steps, existing PTQ methods for diffusion models suffer from distribution mismatch issues at both calibration sample level and reconstruction output level, which makes the performance far from satisfactory, especially in low-bit cases. In this paper, we propose Enhanced Distribution Alignment for Post-Training Quantization of Diffusion Models (EDA-DM) to address the above issues. Specifically, at the calibration sample level, we select calibration samples based on the density and diversity in the latent space, thus facilitating the alignment of their distribution with the overall samples; and at the reconstruction output level, we propose Fine-grained Block Reconstruction, which can align the outputs of the quantized model and the full-precision model at different network granularity. Extensive experiments demonstrate that EDA-DM outperforms the existing post-training quantization frameworks in both unconditional and conditional generation scenarios. At low-bit precision, the quantized models with our method even outperform the full-precision models on most datasets., Comment: 16 pages, 15 figures

We investigate the temperature-dependent production of feebly interacting massive dark matter particle (FIMP DM) within a $Z_2$ model, incorporating two $Z_2$-odd scalar fields. In specific parameter regions, three distinct mechanisms individually dominate the production of the FIMP DM. These mechanisms include the semi-production process, commonly known as the ``exponential growth'' process, the production from pair annihilations of the bath particles, and the three-body decay process. It is crucial to consider the thermal history during the evolution of FIMPs, as it involves multiple phase transitions occurring prior to the freeze-in of dark matter. Consequently, the scalar masses experience thermal variations, leading to a distinctive evolutionary trajectory for FIMPs when compared to scenarios without accounting for the thermal effects. Notably, the unique pattern of FIMP evolution is accompanied by the production of gravitational waves, presenting promising opportunities for detection using forthcoming interferometers., Comment: Published in EPJC

We investigate the inelastic signatures of dark matter-nucleus interactions, explicitly focusing on the ramifications of polarization, dark matter splitting, and the Migdal effect. Direct detection experiments, crucial for testing the existence of dark matter, encounter formidable obstacles such as indomitable neutrino backgrounds and the elusive determination of dark matter spin. To overcome these challenges, we explore the potential of polarized-target dark matter scattering, examining the impact of nonvanishing mass splitting and the role of the Migdal effect in detecting light dark matter. Our analysis demonstrates the valuable utility of the polarized triple-differential event rate as an effective tool for studying inelastic dark matter. It enables us to investigate both angular and energy dependencies, providing valuable insights into the scattering process., Comment: 12 pages, 7 figures, reference updated

Very recently Pulsar Timing Array (PTA) collaborations have independently reported the evidence for a stochastic gravitational-wave background (SGWB), which can unveil the formation of primordial seeds of inhomogeneities in the early universe. With the SGWB parameters inferred from PTAs data, we can make a prediction of the Primordial Black Hole (PBH) clusters from the domain walls of axion-like particles (ALPs). These primordial seeds can naturally provide a solution to the early Active Galactic Nuclei (AGN) formation indicated by James Webb Space Telescope (JWST). Besides, the mass of ALP is also constrained, $m_a \sim 10^{-15}-10^{-14}$ eV, within the reach of upcoming cavity experiments., Comment: 5 pages, 3 figures

The early kinetic decoupling (eKD) effect is an inevitable ingredient in calculating the relic density of dark matter (DM) for various well-motivated scenarios. It appears naturally in forbidden dark matter annihilation, the main focus of this work, which contains fermionic DM and a light singlet scalar that connects the DM and standard model (SM) leptons. The strong suppression of the scattering between DM and SM particles happens quite early in the DM depletion history, where the DM temperature drops away from the thermal equilibrium, $T_\chi < T_{\rm SM}$, leading to the decreased kinetic energy of DM. The forbidden annihilation thus becomes inefficient since small kinetic energy cannot help exceed the annihilation threshold, naturally leading to a larger abundance. To show the eKD discrepancy, we numerically solve the coupled Boltzmann equations that govern the evolution of DM number density and temperature. It is found that eKD significantly affects the DM abundance, resulting in almost an order of magnitude higher than that by the traditional calculation. We also discuss the constraints from experimental searches on the model parameters, where the viable parameter space shrinks when considering the eKD effect., Comment: 10 pages, matches published version

The $\rm{keV}$ scale sterile neutrino was a qualified candidate for dark matter particles in the Dodelson-Widrow mechanism. But the mixing angle, needed to provide enough amount of dark matter, is in contradiction with the astrophysical observations. To alleviate such tension, we introduce an effective interaction, i.e. $g_a (\phi/\Lambda)\partial_{\mu}a \overline{\nu_\alpha}\gamma^{\mu} \gamma_5 \nu_\alpha$, among Standard Model neutrino $\nu_\alpha$, axion $a$, and singlet $\phi$. The axial-vector interaction form is determined by the axion shift symmetry, and the singlet $\phi$ with dynamically varied vacuum expectation value is introduced to reinforce the axial-vector coupling strength and evade the stringent neutrino oscillation constraints. The effective potential generated by the new interaction {could cancel} the SM counterpart, resulting in an {enhanced converting} probability between SM neutrino and sterile neutrino. Hence, the production rate of sterile neutrinos can be substantially enlarged with smaller mixing compared to the DW mechanism., Comment: 6 pages, 2 figures; publish version for EPJC

The supersymmetry (SUSY) breaking mechanism generally predicts the existence of the sgoldstinos, which can play the role of wave-like dark matter. Due to the ubiquitous coupling to the electromagnetic fields, the light scalar sgoldstino dark matter can lead to the variance of the fine-structure constant. With the precise atomic clock data, we find the SUSY breaking scale $\sqrt{F}$ can be probed up to the GUT scale in the sgoldstino mass range of $10^{-22}$ eV $

We study the cosmological inflation and dark matter (DM) in a unified way within a $Z_3$ complex scalar model. The real and imaginary parts of the complex scalar act as the inflaton and DM respectively. The slow-rolling inflation with non-minimal coupling in both the metric and Palatini formalisms can be realized. We examine the whole parameters space by fully considering the theoretical and experimental constraints. We find that in the low-energy scale, the DM relic density and the DM-nucleon direct scattering experiments favor the mixing angle $|\theta| \lesssim 0.25$, the DM mass $m_\chi \gtrsim 80\rm{GeV}$, and the mass of Higgs-like scalar $m_{h_2} \gtrsim 300\rm{GeV}$. In the high-energy scale, after further considering the cosmological constraints of the scalar spectral index and the tensor-to-scalar ratio for the two forms of inflation, the scalar spectral indices are both $\sim 0.965$, the non-minimum coupling coefficients are $\sim 10^4$ and $\sim 10^9$, and the tensor-to-scalar ratios are $\sim 10^{-3}$ and $\lesssim 10^{-11}$ respectively, which suggests that the inflation under the two formalisms can be distinguished by measuring the tensor-to-scalar ratio with higher precision., Comment: 16 pages, 5 figures

By assuming the photosphere located at the outmost edge of the ejecta, Arnett et al. (1980, 1982, 1989) presented the light curves of homologous explosions in supernovae analytically and numerically to include recombination effects. Actually as homologous expansion proceeds, the photosphere recedes deeper into the ejecta. In this situation, the photosphere radius increases at early times and decreases later on which can be described by a simple method proposed by Liu et al. (2018). To study how the photosphere recession effect the luminosity evolution, we impose a boundary condition on the photosphere to determine the spatial and time distribution of the temperature of the ejecta which is clarified to be reasonable. We find that the photosphere recession reduce the luminosity compared with the previous result without the recession, which can be tested with observations of Type-IIP supernovae., Comment: 9 pages, 3 figures

We study a minimal extension of the Standard Model by introducing three right-handed neutrinos and a new scotogenic scalar doublet, in which the mass splittings between neutral and charged components are responsible for the $W$-boson mass newly measured by the CDF collaboration. This model can not only generate non-vanishing Majorana neutrino masses via the interaction of right-handed neutrinos and scotogenic scalars, but also explain the Universe's missing matter in the form of FIMP dark matter. We also study the influence of the mass splitting on the first order electroweak phase transition, and find that it can further enhance the transition strength and thus induce gravitational waves during the phase transition, which may be detected in the forthcoming detectors such as U-DECIGO., Comment: References updated, accepted for publication in Science Bulletin

We study the superheavy dark matter (DM) scenario in an extended $B-L$ model, where one generation of right-handed neutrino $\nu_R$ is the DM candidate. If there is a new lighter sterile neutrino that co-annihilate with the DM candidate, then the annihilation rate is exponentially enhanced, allowing a DM mass much heavier than the Griest-Kamionkowski bound ($\sim10^5$ GeV). We demonstrate that a DM mass $M_{\nu_R}\gtrsim10^{13}$ GeV can be achieved. Although beyond the scale of any traditional DM searching strategy, this scenario is testable via gravitational waves (GWs) emitted by the cosmic strings from the $U(1)_{B-L}$ breaking. Quantitative calculations show that the DM mass $\mathcal{O}(10^9-10^{13}~{\rm GeV})$ can be probed by future GW detectors., Comment: 13 pages + references; 2 figures and 2 tables. To appear on JHEP

Flavor-specific scalar bosons exist in various Standard Model extensions and couple to a single generation of fermions via a global flavor symmetry breaking mechanism. Given this strategy, we propose a MeV flavor-specific scalar model in dimension-$5$ operator series, which explains the muon g-2 anomaly and proton radius puzzle by coupling with the muon and down-quark at the same time. The framework is consistent with the null result of high-intensity searches. Specifically, the supernova constraints for muon couplings become weakened by including the contribution of down-quark interaction. The parameter space for explaining muon $g-2$ discrepancy is available when $10\%$ energy deposition is required in the energy explosion process in the supernova, but this is ruled out by the $1\%$ energy deposition requirement. We also investigate the searches for mediator and dark matter and the resulting constraints on viable parameter space such as nuclear physics constraints, direct detection for light boosted dark matter, and possible CMB constraints. When compared to conventional dark matter production, light dark matter production has two additional modifications: bound state formation and early kinetic equilibrium decoupling. We are now looking into the implications of these effects on the relic density of light dark matter., Comment: Accepted for publication in Sci. China-Phys. Mech. Astron

Domain walls can form after breakdown of a discrete symmetry induced by first-order phase transition, we study the heavy dark matter produced around the temperature of the phase transition that yields the breakdown of a $\mathbb{Z}_{3}$ symmetry. The generated gravitational waves by domain walls decay is found to be able to probed by the Pulsar Timing Arrays, and the future Square Kilometer Array., Comment: 8 pages, 4 figures, comments welcome!

In this work, we study the long-standing experimental anomaly in muon $(g-2)$ and also recent anomalous excess in $K_L\to \pi^0+\nu\bar\nu$ at the J-PARC KOTO experiment with sgoldstino. After supersymmetry breaking, the interactions between quarks and sgoldstino ($s$) make the decays $K\to \pi+s$ sizable through loop diagrams, which affects the measurements of decays $K\to \pi+\mathrm{invisible}$. Furthermore, the couplings between photons and sgoldstino contribute to $\Delta a_\mu$ as well as the bino-slepton contribution. With satisfying all known experimental constraints such as from NA62, E949, E137, Orsay, KTEV and CHARM experiments, these two anomalies can be explained simultaneously. The mass of CP-even sgoldstino is close to the neutral pion mass which does not violate the Grossman-Nir bound. The parameter space can be further tested in future NA62, DUNE experiments, as well as experiments in the LHC., Comment: 17 pages, 4 figures, version accepted by JHEP

Latency in the control loop of adaptive optics (AO) systems can severely limit performance. Under the frozen flow hypothesis linear predictive control techniques can overcome this, however identification and tracking of relevant turbulent parameters (such as wind speeds) is required for such parametric techniques. This can complicate practical implementations and introduce stability issues when encountering variable conditions. Here we present a nonlinear wavefront predictor using a Long Short-Term Memory (LSTM) artificial neural network (ANN) that assumes no prior knowledge of the atmosphere and thus requires no user input. The ANN is designed to predict the open-loop wavefront slope measurements of a Shack-Hartmann wavefront sensor (SH-WFS) one frame in advance to compensate for a single-frame delay in a simulated $7\times7$ single-conjugate adaptive optics (SCAO) system operating at 150 Hz. We describe how the training regime of the LSTM ANN affects prediction performance and show how the performance of the predictor varies under various guide star magnitudes. We show that the prediction remains stable when both wind speed and direction are varying. We then extend our approach to a more realistic two-frame latency system. AO system performance when using the LSTM predictor is enhanced for all simulated conditions with prediction errors within 19.9 to 40.0 nm RMS of a latency-free system operating under the same conditions compared to a bandwidth error of $78.3\pm4.4$ nm RMS.

The concept of $\lambda$-differential operators is a natural generalization of differential operators and difference operators. In this paper, we determine the $\lambda$-differential Lie algebraic structure on the Witt algebra and the Virasoro algebra for invertible $\lambda$. Then we consider several families of modules over the Virasoro algebra with explicit module actions and determine the $\lambda$-differential module structures on them., Comment: 17 pages