Search

3D Gaussians have recently emerged as an efficient representation for novel view synthesis. This work studies its editability with a particular focus on the inpainting task, which aims to supplement an incomplete set of 3D Gaussians with additional points for visually harmonious rendering. Compared to 2D inpainting, the crux of inpainting 3D Gaussians is to figure out the rendering-relevant properties of the introduced points, whose optimization largely benefits from their initial 3D positions. To this end, we propose to guide the point initialization with an image-conditioned depth completion model, which learns to directly restore the depth map based on the observed image. Such a design allows our model to fill in depth values at an aligned scale with the original depth, and also to harness strong generalizability from largescale diffusion prior. Thanks to the more accurate depth completion, our approach, dubbed InFusion, surpasses existing alternatives with sufficiently better fidelity and efficiency under various complex scenarios. We further demonstrate the effectiveness of InFusion with several practical applications, such as inpainting with user-specific texture or with novel object insertion., Comment: Project page: https://johanan528.github.io/Infusion

Non-negative trigonometric polynomials satisfying certain properties are employed when studying a number of aspects of the Riemann zeta function. When establishing zero-free regions in the critical strip, the classical polynomial $3+4\cos(\theta)+\cos(2\theta)$ used by de la Vall\'ee Poussin has since been replaced by more beneficial polynomials with larger degree. The classical polynomial was also employed by Titchmarsh to provide a lower bound on $|\zeta(\sigma+it)|$ when $\sigma>1$. We show that this polynomial is optimal for this purpose., Comment: 8 pages

The development of single-platform qubits, predominant for most of the last few decades, has driven the progress of quantum information technologies but also highlighted the limitations of various platforms. Some inherent issues such as charge/spin noise in materials hinder certain platforms, while increased decoherence upon attempts to scale-up severely impact qubit quality and coupling on others. In addition, a universal solution for coherent information transfer between quantum systems remains lacking. By combining one or more qubit platforms, one could potentially create new hybrid platforms that might alleviate significant issues that current single platform qubits suffer from, and in some cases, even facilitate the conversion of static to flying qubits on the same hybrid platform. While nascent, this is an area of rising importance that could shed new light on robust and scalable qubit development and provide new impetus for research directions. Here, we define the requirements for hybrid systems with artificial atoms in solid state, exemplify them with systems that have been proposed or attempted, and conclude with our outlook for such hybrid quantum systems., Comment: 31 pages, 7 figures

We investigate the neutrino elastic differential cross-section (NDCS) and corresponding mean free path for neutral current scattering in the dense matter of a neutron star. A wide range of observed neutron star (NS) masses is considered, including the presence of $\Lambda$, $\Xi^{-}$, and $\Xi^{0}$ hyperons in the heaviest stars. Their presence significantly decreases the total neutrino mean free path in the heavier stars., Comment: 14 pages, 7 figures, 1 table, PRC published version

The CMS detector will be upgraded for the HL-LHC to include a MIP Timing Detector (MTD). The MTD will consist of barrel and endcap timing layers, BTL and ETL respectively, providing precision timing of charged particles. The BTL sensors are based on LYSO:Ce scintillation crystals coupled to SiPMs with TOFHIR2 ASICs for the front-end readout. A resolution of 30-60 ps for MIP signals at a rate of 2.5 Mhit/s per channel is expected along the HL-LHC lifetime. We present an overview of the TOFHIR2 requirements and design, simulation results and measurements with TOFHIR2 ASICs. The measurements of TOFHIR2 associated to sensor modules were performed in different test setups using internal test pulses or blue and UV laser pulses emulating the signals expected in the experiment. The measurements show a time resolution of 24 ps initially during Beginning of Operation (BoO) and 58 ps at End of Operation (EoO) conditions, matching well the BTL requirements. We also showed that the time resolution is stable up to the highest expected MIP rate. Extensive radiation tests were performed, both with x-rays and heavy ions, showing that TOFHIR2 is not affected by the radiation environment during the experiment lifetime., Comment: Editor: J. Varela

Photoacoustic tomography (PAT) is a rapidly-evolving medical imaging modality that combines optical absorption contrast with ultrasound imaging depth. One challenge in PAT is image reconstruction with inadequate acoustic signals due to limited sensor coverage or due to the density of the transducer array. Such cases call for solving an ill-posed inverse reconstruction problem. In this work, we use score-based diffusion models to solve the inverse problem of reconstructing an image from limited PAT measurements. The proposed approach allows us to incorporate an expressive prior learned by a diffusion model on simulated vessel structures while still being robust to varying transducer sparsity conditions., Comment: 5 pages

Spin ensembles play a pivotal role in various quantum applications such as metrology and simulating many-body physics. Recent research has proposed utilizing spin cat states to encode logical quantum information, with potentially logical lifetimes on the order of seconds via enhanced collective interactions that scale with system size. We investigate the dynamics of spin cat states under inhomogeneous broadening, revealing a phenomenon termed `parity-sensitive inhomogeneous dephasing': odd cat states are significantly more susceptible to inhomogeneous dephasing compared to even cat states due to parity symmetry. Additionally, from a mean-field analysis of the driven-dissipative dynamics, we identify a synchronization phase transition wherein the ensemble becomes completely dephased beyond a critical inhomogeneous linewidth. Our findings shed light on the stability of collective spin states, important for advancing quantum technologies., Comment: 10 pages, 4 figures

Scandium aluminum nitride (ScAlN) has recently emerged as an attractive material for integrated photonics due to its favorable nonlinear optical properties and compatibility with CMOS fabrication. Despite the promising and versatile material properties, it is still an outstanding challenge to realize low-loss photonic circuits on thin-film ScAlN-on-insulator wafers. Here, we present a systematic study on the material quality of sputtered thin-film ScAlN produced in a CMOS-compatible 200 mm line, and an optimized fabrication process to yield 400 nm thick, fully etched waveguides. With surface polishing and annealing, we achieve micro-ring resonators with an intrinsic quality factor as high as $1.47\times 10^5$, corresponding to a propagation loss of 2.4 dB/cm. These results serve as a critical step towards developing future large-scale, low-loss photonic integrated circuits based on ScAlN.

Shock-bubble interactions (SBI) are important across a wide range of physical systems. In inertial confinement fusion, interactions between laser-driven shocks and micro-voids in both ablators and foam targets generate instabilities that are a major obstacle in achieving ignition. Experiments imaging the collapse of such voids at high energy densities (HED) are constrained by spatial and temporal resolution, making simulations a vital tool in understanding these systems. In this study, we benchmark several radiation and thermal transport models in the xRAGE hydrodynamic code against experimental images of a collapsing mesoscale void during the passage of a 300 GPa shock. We also quantitatively examine the role of transport physics in the evolution of the SBI. This allows us to understand the dynamics of the interaction at timescales shorter than experimental imaging framerates. We find that all radiation models examined reproduce empirical shock velocities within experimental error. Radiation transport is found to reduce shock pressures by providing an additional energy pathway in the ablation region, but this effect is small ($\sim$1\% of total shock pressure). Employing a flux-limited Spitzer model for heat conduction, we find that flux limiters between 0.03 and 0.10 produce agreement with experimental velocities, suggesting that the system is well-within the Spitzer regime. Higher heat conduction is found to lower temperatures in the ablated plasma and to prevent secondary shocks at the ablation front, resulting in weaker primary shocks. Finally, we confirm that the SBI-driven instabilities observed in the HED regime are baroclinically driven, as in the low energy case., Comment: 14 pages, 12 figures

Grading short answer questions automatically with interpretable reasoning behind the grading decision is a challenging goal for current transformer approaches. Justification cue detection, in combination with logical reasoners, has shown a promising direction for neuro-symbolic architectures in ASAG. But, one of the main challenges is the requirement of annotated justification cues in the students' responses, which only exist for a few ASAG datasets. To overcome this challenge, we contribute (1) a weakly supervised annotation procedure for justification cues in ASAG datasets, and (2) a neuro-symbolic model for explainable ASAG based on justification cues. Our approach improves upon the RMSE by 0.24 to 0.3 compared to the state-of-the-art on the Short Answer Feedback dataset in a bilingual, multi-domain, and multi-question training setup. This result shows that our approach provides a promising direction for generating high-quality grades and accompanying explanations for future research in ASAG and educational NLP.

Large language models (LLMs) are typically trained on general source data for various domains, but a recent surge in domain-specific LLMs has shown their potential to outperform general-purpose models in domain-specific tasks (e.g., biomedicine). Although domain-specific pre-training enhances efficiency and leads to smaller models, the computational costs of training these LLMs remain high, posing budgeting challenges. We introduce MediSwift, a suite of biomedical LMs that leverage sparse pre-training on domain-specific biomedical text data. By inducing up to 75% weight sparsity during the pre-training phase, MediSwift achieves a 2-2.5x reduction in training FLOPs. Notably, all sparse pre-training was performed on the Cerebras CS-2 system, which is specifically designed to realize the acceleration benefits from unstructured weight sparsity, thereby significantly enhancing the efficiency of the MediSwift models. Through subsequent dense fine-tuning and strategic soft prompting, MediSwift models outperform existing LLMs up to 7B parameters on biomedical tasks, setting new benchmarks w.r.t efficiency-accuracy on tasks such as PubMedQA. Our results show that sparse pre-training, along with dense fine-tuning and soft prompting, offers an effective method for creating high-performing, computationally efficient models in specialized domains., Comment: 14 pages, 2 Figures, 5 Tables (Main Paper) + 3 pages (Supplementary Material). Published at ACL 2024

This study investigates the spatial distribution of inertial particles in turbulent Taylor-Couette flow. Direct numerical simulations are performed using a one-way coupled Eulerian-Lagrangian approach, with a fixed inner wall Reynolds number of 2500 for the carrier flow, while the particle Stokes number varies from 0.034 to 1 for the dispersed phase. We first examine the issue of preferential concentration of particles near the outer wall region. Employing two-dimensional (2D) Voronoi analysis, we observe a pronounced particle clustering with increasing $St$, particularly evident in regions of low fluid velocity. Additionally, we investigate the concentration balance equation, inspired by the work of johnson et al.(2020), to examine particle radial distribution. We discern the predominant sources of influence, namely biased sampling, turbophoresis, and centrifugal effects. Across all cases, centrifugal force emerges as the primary driver, causing particle migration towards the outer wall. Biased sampling predominantly affects smaller inertial particles, driving them towards the inner wall due to sampling within Taylor rolls with inward radial velocity. Conversely, turbophoresis primarily impacts larger inertial particles, inducing migration towards both walls where turbulent intensity is weaker compared to the bulk. With the revealed physics, our work provides a basis for predicting and controlling particle movement and distribution in industrial applications., Comment: 18 pages, 11 figures

The theory of causal emergence (CE) with effective information (EI) posits that complex systems can exhibit CE, where macro-dynamics show stronger causal effects than micro-dynamics. A key challenge of this theory is its dependence on coarse-graining method. In this paper, we introduce a fresh concept of approximate dynamical reversibility and establish a novel framework for CE based on this. By applying singular value decomposition(SVD) to Markov dynamics, we find that the essence of CE lies in the presence of redundancy, represented by the irreversible and correlated information pathways. Therefore, CE can be quantified as the potential maximal efficiency increase for dynamical reversibility or information transmission. We also demonstrate a strong correlation between the approximate dynamical reversibility and EI, establishing an equivalence between the SVD and EI maximization frameworks for quantifying CE, supported by theoretical insights and numerical examples from Boolean networks, cellular automata, and complex networks. Importantly, our SVD-based CE framework is independent of specific coarse-graining techniques and effectively captures the fundamental characteristics of the dynamics., Comment: 55 pages,11 figures

Certain precessing black-hole mergers produce gravitational waves with net circular polarization, understood as an imbalance between right- and left-handed amplitudes. According to the Cosmological Principle, such emission must average to zero across all binary mergers in our Universe to preserve mirror-reflection symmetry at very large scales. We present a new independent gravitational-wave test of this hypothesis. Using a novel observable based on the Chern-Pontryagin pseudo-scalar, we measure the emission of net circular polarization across 47 black-hole mergers recently analyzed by Islam et. al. with a state-of-the art model for precessing black-hole mergers in General Relativity. The average value obtained is consistent with zero. Remarkably, however, we find that at least $82\%$ of the analysed sources must have produced net circular polarization. Of these, GW200129 shows strong evidence for mirror asymmetry, with a Bayes Factor of 12.6 or, equivalently, $93.1\%$ probability. We obtain consistent (although stronger) results of $97.5\%$ and $94.3\%$ respectively using public results on this event from Hannam et. al. and performing our own parameter inference. This finding further implies evidence of astrophysical sources that can spontaneously emit circularly polarized photons by quantum effects. Forthcoming black-hole merger detections will enable stronger constraints on large-scale mirror asymmetry and the Cosmological Principle., Comment: 10 pages, 6 Figures, 3 Appendixes. Version accepted for publication in Physical Review Letters

Tuning language models for dialogue generation has been a prevalent paradigm for building capable dialogue agents. Yet, traditional tuning narrowly views dialogue generation as resembling other language generation tasks, ignoring the role disparities between two speakers and the multi-round interactive process that dialogues ought to be. Such a manner often leads to unsatisfactory chat consistency for the built agent. In this work, we emphasize the interactive, communicative nature of dialogue and argue that it is more feasible to model the speaker roles of agent and user separately, enabling the agent to adhere to its role consistently. With this in mind, we propose an efficient Multi-round Interactive Dialogue Tuning (Midi-Tuning) framework. It models the agent and user individually with two adapters built upon large language models. The adapters make use of respective utterances round by round in alternating order and they are tuned via a round-level memory caching mechanism. Extensive experiments demonstrate that, our framework performs superior to traditional fine-tuning and harbors the tremendous potential for improving dialogue consistency., Comment: Accepted by ACL 2024

This paper aims to analyze errors in the implementation of the Physics-Informed Neural Network (PINN) for solving the Allen--Cahn (AC) and Cahn--Hilliard (CH) partial differential equations (PDEs). The accuracy of PINN is still challenged when dealing with strongly non-linear and higher-order time-varying PDEs. To address this issue, we introduce a stable and bounded self-adaptive weighting scheme known as Residuals-RAE, which ensures fair training and effectively captures the solution. By incorporating this new training loss function, we conduct numerical experiments on 1D and 2D AC and CH systems to validate our theoretical findings. Our theoretical analysis demonstrates that feedforward neural networks with two hidden layers and tanh activation function effectively bound the PINN approximation errors for the solution field, temporal derivative, and nonlinear term of the AC and CH equations by the training loss and number of collocation points.