arXiv subjects

Yunyang Li

Publications and source records attributed to Yunyang Li.

At least 19 recordsLinked to original sources

arXiv2026

Detectors for CLASS-W2: The second 90 GHz telescope of the Cosmology Large Angular Scale Surveyor

The Cosmology Large Angular Scale Surveyor (CLASS) is measuring the Cosmic Microwave Background (CMB) polarization anisotropy on the largest angular scales (>1 degree) to probe the epochs of inflation and reionization. To enhance the CMB mapping speed, we have built, tested, and commissioned in August 2025 a second 90 GHz receiver (CLASS-W2) with a detector focal plane composed of feedhorn-coupled Transition Edge Sensor (TES) bolometers fabricated at NIST-Boulder. The focal plane consists of four modules, each containing 37 feedhorns coupling orthogonal polarizations onto two TES bolometers, for a total of 296 optically sensitive detectors. Laboratory tests show highly uniform TES properties with an array average critical temperature of 184+-3 mK, a thermal conductance of 460+-47 pW/K, and a normal resistance of 7.8+-0.3 mOhms. The detector array has an average band center frequency of 95.2 GHz with 28.3 GHz bandwidth, and achieves a detector yield of 94%. On-sky measurements indicate a mean detector optical load of 3.3 pW, corresponding to an antenna temperature of ~23 K. The array's average beam solid angle is 124 $\mu$sr, with a full width at half maximum of 0.592 degrees, and the end-to-end average optical efficiency is 0.37. We find that high-frequency 'blue-leak' radiation couples directly to the TES bolometer islands; adding a metal-mesh low-pass filter with cutoff frequency of 157 GHz in front of the focal plane suppresses the 'blue-leak' power by 0.9 pW. The four-module array achieves a noise-equivalent temperature of NET= 16 uKrtS. Adding this array has boosted the CLASS 90 GHz mapping speed by 41%.

astro-ph.IM↗

arXiv2026

Uncertainty Modeling for SysML v2

Uncertainty is inherent in modern engineered systems, including cyber-physical systems, autonomous systems, and large-scale software-intensive infrastructures (such as microservice-based systems) operating in dynamic and partially observable environments. The recent publication of Precise Semantics for Uncertainty Modeling (PSUM) by the Object Management Group represents the first standardized specification for uncertainty modeling within the Model-Based Systems Engineering (MBSE) community, providing formally defined semantics for representing and reasoning about uncertainty in models. In parallel, the second version of Systems Modeling Language (SysML v2) was released as the next-generation systems modeling language, offering improved semantic rigor and reusability, yet lacking native constructs aligned with PSUM for first-class uncertainty representation. This paper proposes a systematic extension of SysML v2 that incorporates the PSUM metamodel into its modeling framework. The extension enables explicit specification of indeterminacy sources, structured characterization of uncertainties, and consistent propagation of uncertainty within system models, while preserving conformance with SysML v2 syntax and semantics. We validate the approach through seven case studies. Results demonstrate that the proposed extension (PSUM-SysMLv2) is expressive and applicable for uncertainty-aware MBSE, and potentially enables uncertainty and uncertainty propagation analyses.

cs.SE↗

arXiv2026

Elign: Equivariant Diffusion Model Alignment from Foundational Machine Learning Force Fields

Generative models for 3D molecular conformations must respect Euclidean symmetries and concentrate probability mass on thermodynamically favorable, mechanically stable structures. However, E(3)-equivariant diffusion models often reproduce biases from semi-empirical training data rather than capturing the equilibrium distribution of a high-fidelity Hamiltonian. While physics-based guidance can correct this, it faces two computational bottlenecks: expensive quantum-chemical evaluations (e.g., DFT) and the need to repeat such queries at every sampling step. We present Elign, a post-training framework that amortizes both costs. First, we replace expensive DFT evaluations with a faster, pretrained foundational machine-learning force field (MLFF) to provide physical signals. Second, we eliminate repeated run-time queries by shifting physical steering to the training phase. To achieve the second amortization, we formulate reverse diffusion as a reinforcement learning problem and introduce Force--Energy Disentangled Group Relative Policy Optimization (FED-GRPO) to fine-tune the denoising policy. FED-GRPO includes a potential-based energy reward and a force-based stability reward, which are optimized and group-normalized independently. Experiments show that Elign generates conformations with lower gold-standard DFT energies and forces, while improving stability. Crucially, inference remains as fast as unguided sampling, since no energy evaluations are required during generation.

cs.LG↗

arXiv2026

E2Former-V2: On-the-Fly Equivariant Attention with Linear Activation Memory

Equivariant Graph Neural Networks (EGNNs) have become a widely used approach for modeling 3D atomistic systems. However, mainstream architectures face critical scalability bottlenecks due to the explicit construction of geometric features or dense tensor products on \textit{every} edge. To overcome this, we introduce \textbf{E2Former-V2}, a scalable architecture that integrates algebraic sparsity with hardware-aware execution. We first propose \textbf{E}quivariant \textbf{A}xis-\textbf{A}ligned \textbf{S}parsification (EAAS). EAAS builds on Wigner-$6j$ convolution by exploiting an $\mathrm{SO}(3) \rightarrow \mathrm{SO}(2)$ change of basis to transform computationally expensive dense tensor contractions into efficient, sparse parity re-indexing operations. Building on this representation, we introduce \textbf{On-the-Fly Equivariant Attention}, a fully node-centric mechanism implemented via a custom fused Triton kernel. By eliminating materialized edge tensors and maximizing SRAM utilization, our kernel achieves a \textbf{20$\times$ improvement in TFLOPS} compared to standard implementations. Extensive experiments on the SPICE and OMol25 datasets demonstrate that E2Former-V2 maintains comparable predictive performance while notably accelerating inference. This work demonstrates that large equivariant transformers can be trained efficiently using widely accessible GPU platforms. The code is avalible at https://github.com/IQuestLab/UBio-MolFM/tree/e2formerv2.

cs.LG↗

arXiv2025

Enhancing the Scalability and Applicability of Kohn-Sham Hamiltonians for Molecular Systems

Density Functional Theory (DFT) is a pivotal method within quantum chemistry and materials science, with its core involving the construction and solution of the Kohn-Sham Hamiltonian. Despite its importance, the application of DFT is frequently limited by the substantial computational resources required to construct the Kohn-Sham Hamiltonian. In response to these limitations, current research has employed deep-learning models to efficiently predict molecular and solid Hamiltonians, with roto-translational symmetries encoded in their neural networks. However, the scalability of prior models may be problematic when applied to large molecules, resulting in non-physical predictions of ground-state properties. In this study, we generate a substantially larger training set (PubChemQH) than used previously and use it to create a scalable model for DFT calculations with physical accuracy. For our model, we introduce a loss function derived from physical principles, which we call Wavefunction Alignment Loss (WALoss). WALoss involves performing a basis change on the predicted Hamiltonian to align it with the observed one; thus, the resulting differences can serve as a surrogate for orbital energy differences, allowing models to make better predictions for molecular orbitals and total energies than previously possible. WALoss also substantially accelerates self-consistent-field (SCF) DFT calculations. Here, we show it achieves a reduction in total energy prediction error by a factor of 1347 and an SCF calculation speed-up by a factor of 18%. These substantial improvements set new benchmarks for achieving accurate and applicable predictions in larger molecular systems.

physics.chem-ph↗

arXiv2025

PROTOCALC, A W-band polarized calibrator for CMB Telescopes: application to Simons Observatory and CLASS

Current- and next-generation Cosmic Microwave Background (CMB) experiments will measure polarization anisotropies with unprecedented sensitivities. The need for high precision in these measurements underscores the importance of gaining a comprehensive understanding of instrument properties, with a particular emphasis on the study of the beam properties and, in particular, their polarization characteristics, and the measurement of the polarization angle. In this context, a major challenge lies in the scarcity of millimeter polarized astrophysical sources with sufficient brightness and calibration knowledge to meet the stringent accuracy requirements of future CMB missions. This led to the development of a drone-borne calibration source designed for frequency band centered on approximately 90 GHz band, matching a commonly used channel in ground based CMB measurements. The PROTOtype CALibrator for Cosmology, PROTOCALC, has undergone thorough in-lab testing, and its properties have been subsequently modeled through simulation software integrated into the standard Simons Observatory (SO) analysis pipeline. Moreover, the PROTOCALC system has been tested in the field, having been deployed twice on calibration campaigns with CMB telescopes in the Atacama desert. The data collected constrain the roll angle of the source with a statistical accuracy of $0.045^\circ$.

astro-ph.IM↗

arXiv2025

Efficient and Scalable Density Functional Theory Hamiltonian Prediction through Adaptive Sparsity

Hamiltonian matrix prediction is pivotal in computational chemistry, serving as the foundation for determining a wide range of molecular properties. While SE(3) equivariant graph neural networks have achieved remarkable success in this domain, their substantial computational cost--driven by high-order tensor product (TP) operations--restricts their scalability to large molecular systems with extensive basis sets. To address this challenge, we introduce SPHNet, an efficient and scalable equivariant network, that incorporates adaptive SParsity into Hamiltonian prediction. SPHNet employs two innovative sparse gates to selectively constrain non-critical interaction combinations, significantly reducing tensor product computations while maintaining accuracy. To optimize the sparse representation, we develop a Three-phase Sparsity Scheduler, ensuring stable convergence and achieving high performance at sparsity rates of up to 70%. Extensive evaluations on QH9 and PubchemQH datasets demonstrate that SPHNet achieves state-of-the-art accuracy while providing up to a 7x speedup over existing models. Beyond Hamiltonian prediction, the proposed sparsification techniques also hold significant potential for improving the efficiency and scalability of other SE(3) equivariant networks, further broadening their applicability and impact. Our code can be found at https://github.com/microsoft/SPHNet.

cs.LG↗

arXiv2025

E2Former: An Efficient and Equivariant Transformer with Linear-Scaling Tensor Products

Equivariant Graph Neural Networks (EGNNs) have demonstrated significant success in modeling microscale systems, including those in chemistry, biology and materials science. However, EGNNs face substantial computational challenges due to the high cost of constructing edge features via spherical tensor products, making them impractical for large-scale systems. To address this limitation, we introduce E2Former, an equivariant and efficient transformer architecture that incorporates the Wigner $6j$ convolution (Wigner $6j$ Conv). By shifting the computational burden from edges to nodes, the Wigner $6j$ Conv reduces the complexity from $O(|\mathcal{E}|)$ to $ O(| \mathcal{V}|)$ while preserving both the model's expressive power and rotational equivariance. We show that this approach achieves a 7x-30x speedup compared to conventional $\mathrm{SO}(3)$ convolutions. Furthermore, our empirical results demonstrate that the derived E2Former mitigates the computational challenges of existing approaches without compromising the ability to capture detailed geometric information. This development could suggest a promising direction for scalable and efficient molecular modeling.

cs.LG↗

arXiv2025

A Measurement of the Largest-Scale CMB E-mode Polarization with CLASS

We present measurements of large-scale cosmic microwave background (CMB) E-mode polarization from the Cosmology Large Angular Scale Surveyor (CLASS) 90 GHz data. Using 115 det-yr of observations collected through 2024 with a variable-delay polarization modulator, we achieved a polarization sensitivity of $78\,\mathrm{\mu K\,arcmin}$, comparable to Planck at similar frequencies (100 and 143 GHz). The analysis demonstrates effective mitigation of systematic errors and addresses challenges to large-angular-scale power recovery posed by time-domain filtering in maximum-likelihood map-making. A novel implementation of the pixel-space transfer matrix is introduced, which enables efficient filtering simulations and bias correction in the power spectrum using the quadratic cross-spectrum estimator. Overall, we achieved an unbiased time-domain filtering correction to recover the largest angular scale polarization, with the only power deficit, arising from map-making non-linearity, being characterized as less than $3\%$. Through cross-correlation with Planck, we detected the cosmic reionization at $99.4\%$ significance and measured the reionization optical depth $\tau=0.053^{+0.018}_{-0.019}$, marking the first ground-based attempt at such a measurement. At intermediate angular scales ($\ell>30$), our results, both independently and in cross-correlation with Planck, remain fully consistent with Planck's measurements.

astro-ph.CO↗

arXiv2024

High-Efficiency and Low-Noise Detectors for the Upgraded CLASS 90 GHz Focal Plane

We present the in-lab and on-sky performance for the upgraded 90 GHz focal plane of the Cosmology Large Angular Scale Surveyor (CLASS), which had four of its seven detector wafers updated during the austral winter of 2022. The update aimed to improve the transition-edge-sensor (TES) stability and bias range and to realize the high optical efficiency of the sensor design. Modifications included revised circuit terminations, electrical contact between the TES superconductor and the normal metal providing the bulk of the bolometer's heat capacity, and additional filtering on the TES bias lines. The upgrade was successful: 94% of detectors are stable down to 15% of the normal resistance, providing a wide overlapping range of bias voltages for all TESs on a wafer. The median telescope efficiency improved from $0.42^{+0.15}_{-0.22}$ to $0.60^{+0.10}_{-0.32}$ (68% quantiles). For the four upgraded wafers alone, median telescope efficiency increased to $0.65^{+0.06}_{-0.06}$. Given our efficiency estimate for the receiver optics, this telescope efficiency implies a detector efficiency exceeding $0.90$. The overall noise-equivalent temperature of the 90 GHz focal plane improved from 19 $\mu$K$\sqrt{s}$ to 9.7 $\mu$K$\sqrt{s}$.

astro-ph.IM↗

arXiv2024

Sensitivity-Improved Polarization Maps at 40 GHz with CLASS and WMAP data

Improved polarization measurements at frequencies below 70 GHz with degree-level angular resolution are crucial for advancing our understanding of the Galactic synchrotron radiation and the potential polarized anomalous microwave emission and ultimately benefiting the detection of primordial $B$ modes. In this study, we present sensitivity-improved 40 GHz polarization maps obtained by combining the CLASS 40 GHz and WMAP $Q$-band data through a weighted average in the harmonic domain. The decision to include WMAP $Q$-band data stems from similarities in the bandpasses. Leveraging the accurate large-scale measurements from WMAP $Q$ band and the high-sensitivity information from CLASS 40 GHz band at intermediate scales, the noise level at $\ell\in[30, 100]$ is reduced by a factor of $2-3$ in the map space. A pixel domain analysis of the polarized synchrotron spectral index ($\beta_s$) using WMAP $K$ band and the combined maps (mean and 16/84th percentile across the $\beta_s$ map: $-3.08_{-0.20}^{+0.20}$) reveals a stronger preference for spatial variation (PTE for a uniform $\beta_s$ hypothesis smaller than 0.001) than the results obtained using WMAP $K$ and $Ka$ bands ($-3.08_{-0.14}^{+0.14}$). The cross-power spectra of the combined maps follow the same trend as other low-frequency data, and validation through simulations indicates negligible bias introduced by the combination method (sub-percent level in the power spectra). The products of this work are publicly available on $\mathtt{LAMBDA}$.

astro-ph.CO↗

arXiv2024

A Survey of Generative AI for de novo Drug Design: New Frontiers in Molecule and Protein Generation

Artificial intelligence (AI)-driven methods can vastly improve the historically costly drug design process, with various generative models already in widespread use. Generative models for de novo drug design, in particular, focus on the creation of novel biological compounds entirely from scratch, representing a promising future direction. Rapid development in the field, combined with the inherent complexity of the drug design process, creates a difficult landscape for new researchers to enter. In this survey, we organize de novo drug design into two overarching themes: small molecule and protein generation. Within each theme, we identify a variety of subtasks and applications, highlighting important datasets, benchmarks, and model architectures and comparing the performance of top models. We take a broad approach to AI-driven drug design, allowing for both micro-level comparisons of various methods within each subtask and macro-level observations across different fields. We discuss parallel challenges and approaches between the two applications and highlight future directions for AI-driven de novo drug design as a whole. An organized repository of all covered sources is available at https://github.com/gersteinlab/GenAI4Drug.

q-bio.BM↗

arXiv2023

CLASS Data Pipeline and Maps for 40 GHz Observations through 2022

The Cosmology Large Angular Scale Surveyor (CLASS) is a telescope array that observes the cosmic microwave background over 75\% of the sky from the Atacama Desert, Chile, at frequency bands centered near 40, 90, 150, and 220~GHz. This paper describes the CLASS data pipeline and maps for 40~GHz observations conducted from August 2016 to May 2022. We demonstrate how well the CLASS survey strategy, with rapid ($\sim10\,\mathrm{Hz}$) front-end modulation, recovers the large-scale Galactic polarization signal from the ground: the mapping transfer function recovers $\sim75$\% of $EE$, $BB$, and $VV$ power at $\ell=20$ and $\sim45$\% at $\ell=10$. We present linear and circular polarization maps over 75\% of the sky. Simulations based on the data imply the maps have a white noise level of $110\,\mathrm{μK\, arcmin}$ and correlated noise component rising at low-$\ell$ as $\ell^{-2.2}$. The transfer-function-corrected low-$\ell$ component is comparable to the white noise at the angular knee frequencies of $\ell\approx16$ (linear polarization) and $\ell\approx12$ (circular polarization). Finally, we present simulations of the level at which expected sources of systematic error bias the measurements, finding sub-percent bias for the $Λ\mathrm{CDM}$ $EE$ power spectra. Bias from $E$-to-$B$ leakage due to the data reduction pipeline and polarization angle uncertainty approaches the expected level for an $r=0.01$ $BB$ power spectrum. Improvements to the instrument calibration and the data pipeline will decrease this bias.

astro-ph.CO↗

arXiv2023

CLASS Observations of Atmospheric Cloud Polarization at Millimeter Wavelengths

The dynamic atmosphere imposes challenges to ground-based cosmic microwave background observation, especially for measurements on large angular scales. The hydrometeors in the atmosphere, mostly in the form of clouds, scatter the ambient thermal radiation and are known to be the main linearly polarized source in the atmosphere. This scattering-induced polarization is significantly enhanced for ice clouds due to the alignment of ice crystals under gravity, which are also the most common clouds seen at the millimeter-astronomy sites at high altitudes. This work presents a multifrequency study of cloud polarization observed by the Cosmology Large Angular Scale Surveyor (CLASS) experiment on Cerro Toco in the Atacama Desert of northern Chile, from 2016 to 2022, at the frequency bands centered around 40, 90, 150, and 220 GHz. Using a machine-learning-assisted cloud classifier, we made connections between the transient polarized emission found in all four frequencies with the clouds imaged by monitoring cameras at the observing site. The polarization angles of the cloud events are found to be mostly $90^\circ$ from the local meridian, which is consistent with the presence of horizontally aligned ice crystals. The 90 and 150 GHz polarization data are consistent with a power law with a spectral index of $3.90\pm0.06$, while an excess/deficit of polarization amplitude is found at 40/220 GHz compared with a Rayleigh scattering spectrum. These results are consistent with Rayleigh-scattering-dominated cloud polarization, with possible effects from supercooled water absorption and/or Mie scattering from a population of large cloud particles that contribute to the 220 GHz polarization.

astro-ph.IM↗

arXiv2023

CLASS Angular Power Spectra and Map-Component Analysis for 40 GHz Observations through 2022

Measurement of the largest angular scale ($\ell < 30$) features of the cosmic microwave background (CMB) polarization is a powerful way to constrain the optical depth to reionization and search for the signature of inflation through the detection of primordial $B$-modes. We present an analysis of maps covering 73.6\% of the sky made from the $40\,\mathrm{GHz}$ channel of the Cosmology Large Angular Scale Surveyor (CLASS) from 2016 August to 2022 May. Taking advantage of the measurement stability enabled by front-end polarization modulation and excellent conditions from the Atacama Desert, we show this channel achieves higher sensitivity than the analogous frequencies from satellite measurements in the range $10 < \ell < 100$. Simulations show the CLASS linear (circular) polarization maps have a white noise level of $125 \,(130)\,\mathrm{\mu K\, arcmin}$. We measure the Galaxy-masked $EE$ and $BB$ spectra of diffuse synchrotron radiation and compare to space-based measurements at similar frequencies. In combination with external data, we expand measurements of the spatial variations of the synchrotron spectral energy density (SED) to include new sky regions and measure the diffuse SED in the harmonic domain. We place a new upper limit on a background of circular polarization in the range $5 < \ell < 125$ with the first bin showing $D_\ell < 0.023$ $\mathrm{\mu K^2_{CMB}}$ at 95\% confidence. These results establish a new standard for recovery of the largest-scale CMB polarization from the ground and signal exciting possibilities when the higher sensitivity and higher-frequency CLASS channels are included in the analysis.

astro-ph.CO↗

arXiv2023

Microwave Observations of Venus with CLASS

We report on the disk-averaged absolute brightness temperatures of Venus measured at four microwave frequency bands with the Cosmology Large Angular Scale Surveyor (CLASS). We measure temperatures of 432.3 $\pm$ 2.8 K, 355.6 $\pm$ 1.3 K, 317.9 $\pm$ 1.7 K, and 294.7 $\pm$ 1.9 K for frequency bands centered at 38.8, 93.7, 147.9, and 217.5 GHz, respectively. We do not observe any dependence of the measured brightness temperatures on solar illumination for all four frequency bands. A joint analysis of our measurements with lower frequency Very Large Array (VLA) observations suggests relatively warmer ($\sim$ 7 K higher) mean atmospheric temperatures and lower abundances of microwave continuum absorbers than those inferred from prior radio occultation measurements.

astro-ph.EP↗

arXiv2023

Cosmology Large Angular Scale Surveyor (CLASS): 90 GHz Telescope Pointing, Beam Profile, Window Function, and Polarization Performance

The Cosmology Large Angular Scale Surveyor (CLASS) is a telescope array that observes the cosmic microwave background (CMB) over ~75% of the sky from the Atacama Desert, Chile, at frequency bands centered near 40, 90, 150, and 220 GHz. CLASS measures the large angular scale CMB polarization to constrain the tensor-to-scalar ratio and the optical depth to last scattering. This paper presents the optical characterization of the 90GHz telescope, which has been observing since July 2018. Observations of the Moon establish the pointing while dedicated observations of Jupiter are used for beam calibration. The standard deviations of the pointing error in azimuth, elevation, and boresight angle are 1.3, 2.1, and 2.0 arcminutes, respectively, over the first 3 years of observations. This corresponds to a pointing uncertainty ~7% of the beam's full width at half maximum (FWHM). The effective azimuthally-symmetrized instrument 1D beam estimated at 90 GHz has an FWHM of 0.620+/-0.003 deg and a solid angle of 138.7+/-0.6(stats.)+/-1.1(sys.) usr integrated to a radius of 4 deg. The corresponding beam window function drops to b_ell^2 = 0.93, 0.71, 0.14 at ell = 30, 100, 300, respectively. Far-sidelobes are studied using detector-centered intensity maps of the Moon and measured to be at a level of 10^-3 or below relative to the peak. The polarization angle of Tau A estimated from preliminary survey maps is 149.6+/-0.2(stats.) deg in equatorial coordinates. The instrumental temperature-to-polarization (T-to-P) leakage fraction, inferred from per-detector demodulated Jupiter scan data, has a monopole component at the level of 1.7 x 10^-3, a dipole component with an amplitude of 4.3 x 10^-3, with no evidence of quadrupolar leakage.

astro-ph.IM↗

arXiv2023

Long-Short-Range Message-Passing: A Physics-Informed Framework to Capture Non-Local Interaction for Scalable Molecular Dynamics Simulation

Computational simulation of chemical and biological systems using ab initio molecular dynamics has been a challenge over decades. Researchers have attempted to address the problem with machine learning and fragmentation-based methods. However, the two approaches fail to give a satisfactory description of long-range and many-body interactions, respectively. Inspired by fragmentation-based methods, we propose the Long-Short-Range Message-Passing (LSR-MP) framework as a generalization of the existing equivariant graph neural networks (EGNNs) with the intent to incorporate long-range interactions efficiently and effectively. We apply the LSR-MP framework to the recently proposed ViSNet and demonstrate the state-of-the-art results with up to 40% MAE reduction for molecules in MD22 and Chignolin datasets. Consistent improvements to various EGNNs will also be discussed to illustrate the general applicability and robustness of our LSR-MP framework. The code for our experiments and trained model weights could be found at https://github.com/liyy2/LSR-MP.

physics.chem-ph↗