Search

Large language models (LLMs) have recently demonstrated remarkable success in mathematical reasoning. Despite progress in methods like chain-of-thought prompting and self-consistency sampling, these advances often focus on final correctness without ensuring that the underlying reasoning process is coherent and reliable. This paper introduces Step-KTO, a training framework that combines process-level and outcome-level binary feedback to guide LLMs toward more trustworthy reasoning trajectories. By providing binary evaluations for both the intermediate reasoning steps and the final answer, Step-KTO encourages the model to adhere to logical progressions rather than relying on superficial shortcuts. Our experiments on challenging mathematical benchmarks show that Step-KTO significantly improves both final answer accuracy and the quality of intermediate reasoning steps. For example, on the MATH-500 dataset, Step-KTO achieves a notable improvement in Pass@1 accuracy over strong baselines. These results highlight the promise of integrating stepwise process feedback into LLM training, paving the way toward more interpretable and dependable reasoning capabilities.

While perturbation theories constitute a significant foundation of modern quantum system analysis, extending them from the Hermitian to the non-Hermitian regime remains a non-trivial task. In this work, we generalize the Rayleigh-Schr\"odinger perturbation theory to the non-Hermitian regime by employing a geometric formalism. This framework allows us to compute perturbative corrections to eigenstates and eigenvalues of Hamiltonians iteratively to any order. Furthermore, we observe that the recursion equation for the eigenstates resembles the form of the Girard-Newton formulas, which helps us uncover the general solution to the recursion equation. Moreover, we demonstrate that the perturbation method proposed in this paper reduces to the standard Rayleigh-Schr\"odinger perturbation theory in the Hermitian regime., Comment: 8 pages

Audience feedback is crucial for refining video content, yet it typically comes after publication, limiting creators' ability to make timely adjustments. To bridge this gap, we introduce SimTube, a generative AI system designed to simulate audience feedback in the form of video comments before a video's release. SimTube features a computational pipeline that integrates multimodal data from the video-such as visuals, audio, and metadata-with user personas derived from a broad and diverse corpus of audience demographics, generating varied and contextually relevant feedback. Furthermore, the system's UI allows creators to explore and customize the simulated comments. Through a comprehensive evaluation-comprising quantitative analysis, crowd-sourced assessments, and qualitative user studies-we show that SimTube's generated comments are not only relevant, believable, and diverse but often more detailed and informative than actual audience comments, highlighting its potential to help creators refine their content before release.

Construction of a general-purpose post-recognition error corrector poses a crucial question: how can we most effectively train a model on a large mixture of domain datasets? The answer would lie in learning dataset-specific features and digesting their knowledge in a single model. Previous methods achieve this by having separate correction language models, resulting in a significant increase in parameters. In this work, we present Mixture-of-Experts as a solution, highlighting that MoEs are much more than a scalability tool. We propose a Multi-Task Correction MoE, where we train the experts to become an ``expert'' of speech-to-text, language-to-text and vision-to-text datasets by learning to route each dataset's tokens to its mapped expert. Experiments on the Open ASR Leaderboard show that we explore a new state-of-the-art performance by achieving an average relative $5.0$% WER reduction and substantial improvements in BLEU scores for speech and translation tasks. On zero-shot evaluation, NeKo outperforms GPT-3.5 and Claude-Opus with $15.5$% to $27.6$% relative WER reduction in the Hyporadise benchmark. NeKo performs competitively on grammar and post-OCR correction as a multi-task model., Comment: NeKo work has been done in June 2024. NeKo LMs will be open source on https://huggingface.co/nvidia under the MIT license

The advent of large integral field spectroscopic surveys has found that elliptical galaxies (EGs) can be classified into two classes: the fast rotators (whose kinematics are dominated by rotation) and the slow rotators (which exhibit slow or no rotation pattern). It is often suggested that while the slow rotators typically have boxy isophotal shapes, have a high $\alpha$-to-iron abundance ratio, and are quite massive, the fast rotators often exhibit the opposite properties (that is, having disky isophotes, lower $\alpha$-to-iron ratio, and of typical masses). Whether the EGs consist of two distinct populations (i.e., a dichotomy exists), remains an unsolved issue. To examine the existence of the dichotomy, we used a sample of 1,895 EGs from the SDSS-IV MaNGA survey, and measured robustly the stellar kinematics, isophotal shapes, and [Mg/Fe] ratio. We confirmed the previous finding that the bulk of the EGs are disky (65%) and fast rotators (67%), but found no evidence supporting a dichotomy, based on a principal component analysis. The different classes (boxy/disky and slow/fast rotators) of EGs occupy slightly different loci in the principal component space. This may explain the observed trends that led to the premature support of a dichotomy based on small samples of galaxies., Comment: 27 pages, 19 figures, 5 tables. Submitted to ApJ. Comments are welcome!

Given recent advances in generative AI technology, a key question is how large language models (LLMs) can enhance acoustic modeling tasks using text decoding results from a frozen, pretrained automatic speech recognition (ASR) model. To explore new capabilities in language modeling for speech processing, we introduce the generative speech transcription error correction (GenSEC) challenge. This challenge comprises three post-ASR language modeling tasks: (i) post-ASR transcription correction, (ii) speaker tagging, and (iii) emotion recognition. These tasks aim to emulate future LLM-based agents handling voice-based interfaces while remaining accessible to a broad audience by utilizing open pretrained language models or agent-based APIs. We also discuss insights from baseline evaluations, as well as lessons learned for designing future evaluations., Comment: IEEE SLT 2024. The initial draft version has been done in December 2023. Post-ASR Text Processing and Understanding Community and LlaMA-7B pre-training correction model: https://huggingface.co/GenSEC-LLM/SLT-Task1-Llama2-7b-HyPo-baseline

We present a stacked lensing analysis of 96 galaxy clusters selected by the thermal Sunyaev-Zel'dovich (SZ) effect in maps of the cosmic microwave background (CMB). We select foreground galaxy clusters with a $5\sigma$-level SZ threshold in CMB observations from the Atacama Cosmology Telescope, while we define background source galaxies for the lensing analysis with secure photometric redshift cuts in Year 3 data of the Subaru Hyper Suprime Cam survey. We detect the stacked lensing signal in the range of $0.1 < R\, [h^{-1}\mathrm{Mpc}] < 100$ in each of three cluster redshift bins, $0.092

This paper provides a detailed survey of synthetic data techniques. We first discuss the expected goals of using synthetic data in data augmentation, which can be divided into four parts: 1) Improving Diversity, 2) Data Balancing, 3) Addressing Domain Shift, and 4) Resolving Edge Cases. Synthesizing data are closely related to the prevailing machine learning techniques at the time, therefore, we summarize the domain of synthetic data techniques into four categories: 1) Expert-knowledge, 2) Direct Training, 3) Pre-train then Fine-tune, and 4) Foundation Models without Fine-tuning. Next, we categorize the goals of synthetic data filtering into four types for discussion: 1) Basic Quality, 2) Label Consistency, and 3) Data Distribution. In section 5 of this paper, we also discuss the future directions of synthetic data and state three direction that we believe is important: 1) focus more on quality, 2) the evaluation of synthetic data, and 3) multi-model data augmentation.

This study introduces a hybrid fluid simulation approach that integrates generative diffusion models with physics-based simulations, aiming at reducing the computational costs of flow simulations while still honoring all the physical properties of interest. These simulations enhance our understanding of applications such as assessing hydrogen and CO$_2$ storage efficiency in underground reservoirs. Nevertheless, they are computationally expensive and the presence of nonunique solutions can require multiple simulations within a single geometry. To overcome the computational cost hurdle, we propose a hybrid method that couples generative diffusion models and physics-based modeling. We introduce a system to condition the diffusion model with a geometry of interest, allowing to produce variable fluid saturations in the same geometry. While training the model, we simultaneously generate initial conditions and perform physics-based simulations using these conditions. This integrated approach enables us to receive real-time feedback on a single compute node equipped with both CPUs and GPUs. By efficiently managing these processes within one compute node, we can continuously evaluate performance and stop training when the desired criteria are met. To test our model, we generate realizations in a real Berea sandstone fracture which shows that our technique is up to 4.4 times faster than commonly used flow simulation initializations.

We present cosmological constraints using the abundance of weak-lensing shear-selected galaxy clusters in the Hyper Suprime-Cam (HSC) Subaru Strategic Program. The clusters are selected on the mass maps constructed using the three-year (Y3) weak-lensing data with an area of $\approx500~$deg$^2$, resulting in a sample size of $129$ clusters with high signal-to-noise ratios $\nu$ of $\nu\geq4.7$. Owing to the deep, wide-field, and uniform imaging of the HSC survey, this is by far the largest sample of shear-selected clusters, in which the selection solely depends on gravity and is free from any assumptions about the dynamical state. Informed by the optical counterparts, the shear-selected clusters span a redshift range of $z\lesssim0.7$ with a median of $z\approx0.3$. The lensing sources are securely selected at $z\gtrsim0.7$ with a median of $z\approx1.3$, leading to nearly zero cluster member contamination. We carefully account for (1) the bias in the photometric redshift of sources, (2) the bias and scatter in the weak-lensing mass using a simulation-based calibration, and (3) the measurement uncertainty that is directly estimated on the mass maps using an injection-based method developed in a companion paper (Chen et al. submitted). In a blind analysis, the fully marginalized posteriors of the cosmological parameters are obtained as $\Omega_{\mathrm{m}} = 0.50^{+0.28}_{-0.24}$, $\sigma_8 = 0.685^{+0.161}_{-0.088}$, $\hat{S}_{8}\equiv\sigma_8\left(\Omega_{\mathrm{m}}/0.3\right)^{0.25} = 0.835^{+0.041}_{-0.044}$, and $\sigma_8\left(\Omega_{\mathrm{m}}/0.3\right)^{0.5} = 0.993^{+0.084}_{-0.126}$ in a flat $\Lambda$CDM model. We compare our cosmological constraints with other studies, including those based on cluster abundances, galaxy-galaxy lensing and clustering, and Cosmic Microwave Background observed by $Planck$, and find good agreement at levels of $\lesssim2\sigma$. [abridged], Comment: Accepted for publication in the Open Journal of Astrophysics. Main results are Figures 11 and 12. The cluster catalog and the cosmological chain are accessible at https://github.com/inonchiu/hsc_shear_selected_clusters

We present the first step toward deriving cosmological constraints through the abundances of galaxy clusters selected in a $510\,\mathrm{deg}^2$ weak-lensing aperture mass map, constructed with the Year-Three shear catalog from the Hyper Suprime-Cam Subaru Strategic Program. We adopt a conservative source galaxy selection to construct a sample of $129$ weak-lensing peaks with a signal-to-noise ratio above $4.7$. We use semi-analytical injection simulations to derive the selection function and the mass--observable relation of our sample. These results take into account complicated uncertainties associated with weak-lensing measurements, such as the non-uniform survey depth and the complex survey geometry, projection effects from uncorrelated large-scale structures, and the intrinsic alignment of source galaxies. We also propose a novel modeling framework to make parts of the mass--observable relation insensitive to assumed cosmological parameters. Such a framework not only offers a great computational advantage to cosmological studies, but can also benefit future astrophysical studies using shear-selected clusters. Our results are an important step toward utilizing these cluster samples that are constructed nearly independent of any baryonic assumptions in upcoming deep-and-wide lensing surveys from the Vera Rubin Observatory, Euclid, and the Nancy Grace Roman Space Telescope., Comment: 18 pages, 9 figures; accepted for publication in the Open Journal of Astrophysics. The cluster catalog of the cosmological sample (signal-to-noise ratio above $4.7$) is accessible at https://github.com/inonchiu/hsc_shear_selected_clusters. See companion paper Chiu et al (arXiv:2406.11970) for cosmological constraints from this cluster sample

LLMs have gotten attention across various research domains due to their exceptional performance on a wide range of complex tasks. Therefore, refined methods to evaluate the capabilities of LLMs are needed to determine the tasks and responsibility they should undertake. Our study mainly discussed how LLMs, as useful tools, should be effectively assessed. We proposed the two-stage framework: from ``core ability'' to ``agent'', clearly explaining how LLMs can be applied based on their specific capabilities, along with the evaluation methods in each stage. Core ability refers to the capabilities that LLMs need in order to generate high-quality natural language texts. After confirming LLMs possess core ability, they can solve real-world and complex tasks as agent. In the "core ability" stage, we discussed the reasoning ability, societal impact, and domain knowledge of LLMs. In the ``agent'' stage, we demonstrated embodied action, planning, and tool learning of LLMs agent applications. Finally, we examined the challenges currently confronting the evaluation methods for LLMs, as well as the directions for future development.

We present the Liouville Flow Importance Sampler (LFIS), an innovative flow-based model for generating samples from unnormalized density functions. LFIS learns a time-dependent velocity field that deterministically transports samples from a simple initial distribution to a complex target distribution, guided by a prescribed path of annealed distributions. The training of LFIS utilizes a unique method that enforces the structure of a derived partial differential equation to neural networks modeling velocity fields. By considering the neural velocity field as an importance sampler, sample weights can be computed through accumulating errors along the sample trajectories driven by neural velocity fields, ensuring unbiased and consistent estimation of statistical quantities. We demonstrate the effectiveness of LFIS through its application to a range of benchmark problems, on many of which LFIS achieved state-of-the-art performance., Comment: 25 pages, 7 figures, 15 tables. Submitted to and accepted by the 41th International Conference on Machine Learning (Vienna, Austria)

The evaluation of large language models (LLMs) has drawn substantial attention in the field recently. This work focuses on evaluating LLMs in a Chinese context, specifically, for Traditional Chinese which has been largely underrepresented in existing benchmarks. We present TMLU, a holistic evaluation suit tailored for assessing the advanced knowledge and reasoning capability in LLMs, under the context of Taiwanese Mandarin. TMLU consists of an array of 37 subjects across social science, STEM, humanities, Taiwan-specific content, and others, ranging from middle school to professional levels. In addition, we curate chain-of-thought-like few-shot explanations for each subject to facilitate the evaluation of complex reasoning skills. To establish a comprehensive baseline, we conduct extensive experiments and analysis on 24 advanced LLMs. The results suggest that Chinese open-weight models demonstrate inferior performance comparing to multilingual proprietary ones, and open-weight models tailored for Taiwanese Mandarin lag behind the Simplified-Chinese counterparts. The findings indicate great headrooms for improvement, and emphasize the goal of TMLU to foster the development of localized Taiwanese-Mandarin LLMs. We release the benchmark and evaluation scripts for the community to promote future research., Comment: Preprint. Under review

Dislocation mobility, which dictates the response of dislocations to an applied stress, is a fundamental property of crystalline materials that governs the evolution of plastic deformation. Traditional approaches for deriving mobility laws rely on phenomenological models of the underlying physics, whose free parameters are in turn fitted to a small number of intuition-driven atomic scale simulations under varying conditions of temperature and stress. This tedious and time-consuming approach becomes particularly cumbersome for materials with complex dependencies on stress, temperature, and local environment, such as body-centered cubic crystals (BCC) metals and alloys. In this paper, we present a novel, uncertainty quantification-driven active learning paradigm for learning dislocation mobility laws from automated high-throughput large-scale molecular dynamics simulations, using Graph Neural Networks (GNN) with a physics-informed architecture. We demonstrate that this Physics-informed Graph Neural Network (PI-GNN) framework captures the underlying physics more accurately compared to existing phenomenological mobility laws in BCC metals.

Superclusters, encompassing environments across a wide range of overdensities, can be regarded as unique laboratories for studying galaxy evolution. Although numerous supercluster catalogs have been published, none of them goes beyond redshift $z=0.7$. In this work, we adopt a physically motivated supercluster definition, requiring that superclusters should eventually collapse even in the presence of dark energy. Applying a friends-of-friends (FoF) algorithm to the CAMIRA cluster sample constructed using the Subaru Hyper Suprime-Cam survey data, we have conducted the first systematic search for superclusters at $z=0.5-1.0$ and identified 673 supercluster candidates over an area of 1027 deg$^2$. The FoF algorithm is calibrated by evolving $N$-body simulations to the far future to ensure high purity. We found that these high-$z$ superclusters are mainly composed of $2-4$ clusters, suggesting the limit of gravitationally bound structures in the younger Universe. In addition, we studied the properties of the clusters and brightest cluster galaxies (BCGs) residing in different large-scale environments. We found that clusters associated with superclusters are typically richer, but no apparent dependence of the BCG properties on large-scale structures is found. We also compared the abundance of observed superclusters with mock superclusters extracted from halo light cones, finding that photometric redshift uncertainty is a limiting factor in the performance of superclusters detection., Comment: Accepted by ApJ. 36 pages, 26 figures, 7 tables

(Reduced) Using the recent cosmological high-resolution zoom-in simulations, NewHorizon and Galactica, in which the evolution of black hole spin is followed on the fly, we have tracked the cosmic history of a hundred of black holes (BHs) with a mass greater than 2x10^4 Ms. For each of them, we have studied the variations of the three dimensional angle (Psi) subtended between the BH spins and the angular momentum vectors of their host galaxies. The analysis of the individual evolution of the most massive BHs suggests that they are generally passing by three different regimes. First, for a short period after their birth, low mass BHs (<3x10^4 Ms) are rapidly spun up by gas accretion and their spin tends to be aligned with their host galaxy spin. Then follows a second phase in which the accretion of gas onto low mass BHs (<10^5 Ms) is quite chaotic and inefficient, reflecting the complex and disturbed morphologies of forming proto-galaxies at high redshifts. The variations of Psi are rather erratic during this phase and are mainly driven by the rapid changes of the direction of the galaxy angular momentum. Then, in a third and long phase, BHs are generally well settled in the center of galaxies around which the gas accretion becomes much more coherent (>10^5 Ms). In this case, the BH spins tend to be well aligned with the angular momentum of their host galaxy and this configuration is generally stable even though BH merger episodes can temporally induce misalignment. We have also derived the distributions of cos(Psi) at different redshifts and found that BHs and galaxy spins are generally aligned. Finally, based on a Monte Carlo method, we also predict statistics for the 2-d projected spin-orbit angles lambda. In particular, the distribution of lambda traces well the alignment tendency in the 3-d analysis. Such predictions provide an interesting background for future observational analyses., Comment: Accepted for publication in A&A (12/03/2024)

Pore-scale simulations accurately describe transport properties of fluids in the subsurface. These simulations enhance our understanding of applications such as assessing hydrogen storage efficiency and forecasting CO$_2$ sequestration processes in underground reservoirs. Nevertheless, they are computationally expensive due to their mesoscopic nature. In addition, their stationary solutions are not guaranteed to be unique, so multiple runs with different initial conditions must be performed to ensure sufficient sample coverage. These factors complicate the task of obtaining representative and reliable forecasts. To overcome the high computational cost hurdle, we propose a hybrid method that couples generative diffusion models and physics-based modeling. Upon training a generative model, we synthesize samples that serve as the initial conditions for physics-based simulations. We measure the relaxation time (to stationary solutions) of the simulations, which serves as a validation metric and early-stopping criterion. Our numerical experiments revealed that the hybrid method exhibits a speed-up of up to 8.2 times compared to commonly used initialization methods. This finding offers compelling initial support that the proposed diffusion model-based hybrid scheme has potentials to significantly decrease the time required for convergence of numerical simulations without compromising the physical robustness.

We introduce the Mori-Zwanzig (MZ) Modal Decomposition (MZMD), a novel technique for performing modal analysis of large scale spatio-temporal structures in complex dynamical systems, and show that it represents an efficient generalization of Dynamic Mode Decomposition (DMD). The MZ formalism provides a mathematical framework for constructing non-Markovian reduced-order models of resolved variables from high-dimensional dynamical systems, incorporating the effects of unresolved dynamics through the memory kernel and orthogonal dynamics. We present a formulation and analysis of the modes and spectrum from MZMD and compare it to DMD when applied to a complex flow: a Direct Numerical Simulation (DNS) data-set of laminar-turbulent boundary-layer transition flow over a flared cone at Mach 6. We show that the addition of memory terms by MZMD improves the resolution of spatio-temporal structures within the transitional/turbulent regime, which contains features that arise due to nonlinear mechanisms, such as the generation of the so-called "hot" streaks on the surface of the flared cone. As a result, compared to DMD, MZMD improves future state prediction accuracy, while requiring nearly the same computational cost.

Galaxies play a key role in our endeavor to understand how structure formation proceeds in the Universe. For any precision study of cosmology or galaxy formation, there is a strong demand for huge sets of realistic mock galaxy catalogs, spanning cosmologically significant volumes. For such a daunting task, methods that can produce a direct mapping between dark matter halos from dark matter-only simulations and galaxies are strongly preferred, as producing mocks from full-fledged hydrodynamical simulations or semi-analytical models is too expensive. Here we present a Graph Neural Network-based model that is able to accurately predict key properties of galaxies such as stellar mass, $g-r$ color, star formation rate, gas mass, stellar metallicity, and gas metallicity, purely from dark matter properties extracted from halos along the full assembly history of the galaxies. Tests based on the TNG300 simulation of the IllustrisTNG project show that our model can recover the baryonic properties of galaxies to high accuracy, over a wide redshift range ($z = 0-5$), for all galaxies with stellar masses more massive than $10^9\,M_\odot$ and their progenitors, with strong improvements over the state-of-the-art methods. We further show that our method makes substantial strides toward providing an understanding of the implications of the IllustrisTNG galaxy formation model., Comment: 18 pages, 7 figures, 3 tables, 4 pages of appendices. Submitted to ApJ

We present a protocluster search covering $z\sim3$ to $z\sim5$ based on the combination of the Hyper SuprimeCam Subaru Strategic Programme and the CFHT Large Area $U$-band Deep Survey. We identify about 30 protocluster candidates per unit redshift over the $\sim25\,\mathrm{deg^2}$ area of the Deep/Ultra-Deep layer. Protocluster candidates are selected as regions with a significantly enhanced surface density of dropout galaxies. With this large sample, we characterise the properties of their individual member galaxies. We compare the number counts of dropout galaxies in protocluster candidates with that of coeval field galaxies. Rest-frame UV bright galaxies are over-abundant in protocluster candidates, a trend seen across the full redshift range studied. We do not find evidence for their spatial distribution within protocluster candidates to be distinct from their fainter counterparts, nor for their UV colour to be different from that of field galaxies with the same brightness. Cosmological simulations predict this bright-end excess, with the main cause being a richer population of massive galaxies, with only a minor contribution from an enhancement in star formation activity (and therefore UV emission) at fixed mass. $U$-to-$K$ SED modelling of our observed samples supports this interpretation. This environmental differentiation in number counts is already in place at $z\sim5$, with no significant redshift dependence over the range in lookback times probed. These observational results and model predictions suggest that the cosmic clock is ahead in high-density environments., Comment: 17 pages, 13 figures, accepted for publication in MNRAS

Understanding, predicting and controlling laminar-turbulent boundary-layer transition is crucial for the next generation aircraft design. However, in real flight experiments, or wind tunnel tests, often only sparse sensor measurements can be collected at fixed locations. Thus, in developing reduced models for predicting and controlling the flow at the sensor locations, the main challenge is in accounting for how the surrounding field of unobserved variables interacts with the observed variables at the fixed sensor locations. This makes the Mori-Zwanzig (MZ) formalism a natural choice, as it results in the Generalized Langevin Equations which provides a framework for constructing non-Markovian reduced-order models that includes the effects the unresolved variables have on the resolved variables. These effects are captured in the so called memory kernel and orthogonal dynamics. In this work, we explore the data-driven MZ formulations to two boundary layer flows obtained from DNS data; a low speed incompressible flow; and a high speed compressible flow over a flared cone at Mach 6. An array of "sensors" are placed near the surface of the solid boundary, and the MZ operators are learned and the predictions are compared to the Extended Dynamic Mode Decomposition (EDMD), both using delay embedded coordinates. Further comparisons are made with Long Short-Term Memory (LSTM) and a regression based projection framework using neural networks for the MZ operators. First we compare the effects of including delay embedded coordinates with EDMD and Mori based MZ and provide evidence that using both memory and delay embedded coordinates minimizes generalization errors on the relevant time scales. Next, we provide numerical evidence that the data-driven regression based projection MZ model performs best with respect to the prediction accuracy (minimum generalization error) on the relevant time scales., Comment: AIAA-Aviation 2023

Services of personalized TTS systems for the Mandarin-speaking speech impaired are rarely mentioned. Taiwan started the VoiceBanking project in 2020, aiming to build a complete set of services to deliver personalized Mandarin TTS systems to amyotrophic lateral sclerosis patients. This paper reports the corpus design, corpus recording, data purging and correction for the corpus, and evaluations of the developed personalized TTS systems, for the VoiceBanking project. The developed corpus is named after the VoiceBank-2023 speech corpus because of its release year. The corpus contains 29.78 hours of utterances with prompts of short paragraphs and common phrases spoken by 111 native Mandarin speakers. The corpus is labeled with information about gender, degree of speech impairment, types of users, transcription, SNRs, and speaking rates. The VoiceBank-2023 is available by request for non-commercial use and welcomes all parties to join the VoiceBanking project to improve the services for the speech impaired., Comment: submitted to 26th International Conference of the ORIENTAL-COCOSDA

We study an approach to obtaining the exact formal solution of the 2-species Lotka-Volterra equation based on combinatorics and generating functions. By employing a combination of Carleman linearization and Mori-Zwanzig reduction techniques, we transform the nonlinear equations into a linear system, allowing for the derivation of a formal solution. The Mori-Zwanzig reduction reduces to an expansion which we show can be interpreted as a directed and weighted lattice path walk, which we use to obtain a representation of the system dynamics as walks of fixed length. The exact solution is then shown to be dependent on the generator of weighted walks. We show that the generator can be obtained by the solution of PDE which in turn is equivalent to a particular Koopman evolution of nonlinear observables., Comment: 27 single column pages+epsilon, 3 figures

It has become common to perform kinetic analysis using approximate Koopman operators that transforms high-dimensional time series of observables into ranked dynamical modes. Key to a practical success of the approach is the identification of a set of observables which form a good basis in which to expand the slow relaxation modes. Good observables are, however, difficult to identify {\em a priori} and sub-optimal choices can lead to significant underestimations of characteristic timescales. Leveraging the representation of slow dynamics in terms of Hidden Markov Model (HMM), we propose a simple and computationally efficient clustering procedure to infer surrogate observables that form a good basis for slow modes. We apply the approach to an analytically solvable model system, as well as on three protein systems of different complexities. We consistently demonstrate that the inferred indicator functions can significantly improve the estimation of the leading eigenvalues of the Koopman operators and correctly identify key states and transition timescales of stochastic systems, even when good observables are not known {\em a priori}., Comment: 28 pages, 6 figures

We propose LLM-Eval, a unified multi-dimensional automatic evaluation method for open-domain conversations with large language models (LLMs). Existing evaluation methods often rely on human annotations, ground-truth responses, or multiple LLM prompts, which can be expensive and time-consuming. To address these issues, we design a single prompt-based evaluation method that leverages a unified evaluation schema to cover multiple dimensions of conversation quality in a single model call. We extensively evaluate the performance of LLM-Eval on various benchmark datasets, demonstrating its effectiveness, efficiency, and adaptability compared to state-of-the-art evaluation methods. Our analysis also highlights the importance of choosing suitable LLMs and decoding strategies for accurate evaluation results. LLM-Eval offers a versatile and robust solution for evaluating open-domain conversation systems, streamlining the evaluation process and providing consistent performance across diverse scenarios., Comment: Accepted at 5th NLP4ConvAI

We present new Spitzer Infrared Array Camera (IRAC) 3.6 and 4.5{\mu}m mosaics of three fields, E-COSMOS, DEEP2-F3, and ELAIS-N1. Our mosaics include both new IRAC observations as well as re-processed archival data in these fields. These fields are part of the HSC-Deep grizy survey and have a wealth of additional ancillary data. The addition of these new IRAC mosaics is critical in allowing for improved photometric redshifts and stellar population parameters at cosmic noon and earlier epochs. The total area mapped by this work is {\sim} 17 deg2 with a mean integration time of {\sim}1200s, providing a median 5{\sigma} depth of 23.7(23.3) at 3.6(4.5){\mu}m in AB. We perform SExtractor photometry both on the combined mosaics as well as the single-epoch mosaics taken {\sim}6 months apart. The resultant IRAC number counts show good agreement with previous studies. In combination with the wealth of existing and upcoming spectro-photometric data in these fields, our IRAC mosaics will enable a wide range of galactic evolution and AGN studies. With that goal in mind, we make the combined IRAC mosaics and coverage maps of these three fields publicly available. counts show good agreement with previous studies., Comment: accepted for publication in AJ

Typical generative diffusion models rely on a Gaussian diffusion process for training the backward transformations, which can then be used to generate samples from Gaussian noise. However, real world data often takes place in discrete-state spaces, including many scientific applications. Here, we develop a theoretical formulation for arbitrary discrete-state Markov processes in the forward diffusion process using exact (as opposed to variational) analysis. We relate the theory to the existing continuous-state Gaussian diffusion as well as other approaches to discrete diffusion, and identify the corresponding reverse-time stochastic process and score function in the continuous-time setting, and the reverse-time mapping in the discrete-time setting. As an example of this framework, we introduce ``Blackout Diffusion'', which learns to produce samples from an empty image instead of from noise. Numerical experiments on the CIFAR-10, Binarized MNIST, and CelebA datasets confirm the feasibility of our approach. Generalizing from specific (Gaussian) forward processes to discrete-state processes without a variational approximation sheds light on how to interpret diffusion models, which we discuss., Comment: 29 pages, 13 figures, 2 tables. Accepted by the 40th International Conference on Machine Learning, Honolulu, Hawaii, USA

We report CFHT/SITELLE imaging Fourier Transform Spectrograph observations of the Brightest Cluster Galaxy (BCG) of galaxy cluster Abell 2390 at z=0.228. The BCG displays a prominent cone of emission in H$\alpha$, H$\beta$, [NII], and [OII] to the North-West with PA = 42$^o$, 4.4 arcsec in length (15.9 kpc), which is associated with elongated and asymmetric Chandra soft X-ray emission. The H$\alpha$ flux map also contains a "hook" of H$\alpha$ and [NII] emission resulting in a broadened northern edge to the cone. Using SITELLE/LUCI software we extract emission line flux, velocity, velocity dispersion, and continuum maps, and utilize them to derive flux ratio maps to determine ionization mechanisms and dynamical information in the BCG's emission line region. The Baldwin-Phillips-Terlevich diagnostics on the BCG cone indicate a composite ionization origin of photoionization due to star formation and shock. Strong LINER-like emission is seen in the nuclear region which hosts an AGN. As Abell 2390 is a cool-core cluster, we suggest that the cooling flow is falling onto the central BCG and interacting with the central AGN. The AGN produces jets that inflate "bubbles" of plasma in the ICM, as is often observed in local galaxy clusters. Furthermore, combining signs of AGN activities from radio, optical emission line and X-ray data over a large range of physical scale, we find evidence for three possible episodes of AGN activity in different epochs associated with the Abell 2390 BCG., Comment: 15 pages, 13 figures, accepted by MNRAS

This work focuses on in-context data augmentation for intent detection. Having found that augmentation via in-context prompting of large pre-trained language models (PLMs) alone does not improve performance, we introduce a novel approach based on PLMs and pointwise V-information (PVI), a metric that can measure the usefulness of a datapoint for training a model. Our method first fine-tunes a PLM on a small seed of training data and then synthesizes new datapoints - utterances that correspond to given intents. It then employs intent-aware filtering, based on PVI, to remove datapoints that are not helpful to the downstream intent classifier. Our method is thus able to leverage the expressive power of large language models to produce diverse training data. Empirical results demonstrate that our method can produce synthetic training data that achieve state-of-the-art performance on three challenging intent detection datasets under few-shot settings (1.28% absolute improvement in 5-shot and 1.18% absolute in 10-shot, on average) and perform on par with the state-of-the-art in full-shot settings (within 0.01% absolute, on average)., Comment: Accepted at EACL 2023

In this work, we apply, for the first time to spatially inhomogeneous flows, a recently developed data-driven learning algorithm of Mori-Zwanzig (MZ) operators, which is based on a generalized Koopman's description of dynamical systems. The MZ formalism provides a mathematically exact procedure for constructing non-Markovian reduced-order models of resolved variables from high-dimensional dynamical systems, where the effects due to the unresolved dynamics are captured in the memory kernel and orthogonal dynamics. The algorithm developed in this work applies Mori's linear projection operator and an SVD based compression to the selection of the resolved variables (equivalently, a low rank approximation of the two time covariance matrices). We show that this MZ decomposition not only identifies the same spatio-temporal structures found by DMD, but it can also be used to extract spatio-temporal structures of the hysteresis effects present in the memory kernels. We perform an analysis of these structures in the context of a laminar-turbulent boundary-layer transition flow over a flared cone at Mach 6, and show the dynamical relevance of the memory kernels. Additionally, by including these memory terms learned in our data-driven MZ approach, we show improvement in prediction accuracy over DMD at the same level of truncation and at a similar computational cost. Furthermore, an analysis of the spatio-temporal structures of the MZ operators shows identifiable structures associated with the nonlinear generation of the so-called "hot" streaks on the surface of the flared code, which have previously been observed in experiments and direct numerical simulations., Comment: Published in AIAA Scitech 2023 Conference

Subhalo abundance matching (SHAM) has played an important role in improving our understanding of how galaxies populate their host dark matter halos. In essence, the SHAM framework is to find a dark matter halo property that best correlates with an attribute of galaxies, such as stellar mass. The peak value of the maximum circular velocity ($V_{\rm max}$) a halo/subhalo has ever attained throughout its lifetime, $V_{\rm peak}$, has been a popular choice for SHAM. A recent study by Tonnesen & Ostriker (2021) suggested that quantity $\phi$, which combines the present-day $V_{\rm max}$ and the peak value of halo dark matter mass, performs better in predicting stellar mass than $V_{\rm peak}$. Inspired by their approach, in this work, we find that further improvement can be achieved by a quantity $\psi_5$ that combines the 90th percentile of $V_{\rm max}$ a halo/subhalo has ever achieved with the 60th percentile of the dark matter halo time variation rate. Tests based on the simulation IllustrisTNG300 show that our new SHAM scheme, with just three free parameters, can improve the stellar mass prediction and mass-dependent clustering by 15% and 16% from $\phi$, respectively, over the redshift range $z=0-2$., Comment: 16 pages, 8 figures, 3 tables, Accepted by The Astrophysical Journal

We propose a novel subhalo abundance matching (SHAM) model that uses the virial mass of the main progenitor of each (sub)halo $M_{\rm prog}$ as a proxy of the galaxy stellar mass $M_*$ at the time of observation. This $M_{\rm prog}$ model predicts the two-point correlation functions depending on the choice of the epoch $z_{\rm prog}$ at which $M_\mathrm{prog}$ is quoted. With $z_{\rm prog}$ as a fitting parameter, we apply the $M_{\rm prog}$ model to the angular correlation functions measured with varying stellar mass thresholds from $M_{*,~{\rm lim}}/(h^{-2}M_\odot)=10^{11}$ to $10^{8.6}$ using a sample of galaxies at $z\simeq0.4$ from the Subaru Hyper Suprime-Cam survey. The $M_{\rm prog}$ model can reproduce the observations very well over $10~h^{-1}{\rm kpc}\textrm{--}10~h^{-1}{\rm Mpc}$. We find that, for the samples of $10^{9.2}\leq M_{*,~{\rm lim}}/(h^{-2}M_\odot)\leq10^{10.2}$, the correlation functions predicted by the widely-used $V_{\rm peak}$ model lack amplitudes at $\lesssim1~h^{-1}{\rm Mpc}$, suggesting that $M_{\rm prog}$ is a better proxy of the galaxy stellar mass than conventional $V_{\rm peak}$. The $z_{\rm prog}$ parameter is highest ($z_{\rm prog}\simeq3$) for intermediate mass galaxies at $M_*\simeq10^{9.9}~h^{-2}M_\odot$, and becomes smaller down to $z_\mathrm{prog}\simeq1$ for both lower- and higher-mass galaxies. We interpret these trends as reflecting the downsizing in the in-situ star formation in lower-mass galaxies and the larger contribution of the ex-situ stellar mass growth in higher-mass galaxies., Comment: 12 pages, 8 figures, accepted for publication in MNRAS, the angular correlation function data of the HSC galaxies are available on https://sites.google.com/view/smasaki/works-%E7%A0%94%E7%A9%B6%E6%A5%AD%E7%B8%BE?authuser=0

While rich, open-domain textual data are generally available and may include interesting phenomena (humor, sarcasm, empathy, etc.) most are designed for language processing tasks, and are usually in a non-conversational format. In this work, we take a step towards automatically generating conversational data using Generative Conversational Networks, aiming to benefit from the breadth of available language and knowledge data, and train open domain social conversational agents. We evaluate our approach on conversations with and without knowledge on the Topical Chat dataset using automatic metrics and human evaluators. Our results show that for conversations without knowledge grounding, GCN can generalize from the seed data, producing novel conversations that are less relevant but more engaging and for knowledge-grounded conversations, it can produce more knowledge-focused, fluent, and engaging conversations. Specifically, we show that for open-domain conversations with 10\% of seed data, our approach performs close to the baseline that uses 100% of the data, while for knowledge-grounded conversations, it achieves the same using only 1% of the data, on human ratings of engagingness, fluency, and relevance., Comment: Accepted at SIGDial 2022

We study spatially resolved scaling relations among stars, dust, and gas in ten nearby spiral galaxies. In a preceding paper Abdurro'uf et al. (2022), we have derived spatially resolved properties of the stellar population and dust by panchromatic spectral energy distribution (SED) fitting using piXedfit. Now, we investigate resolved star formation ($\Sigma_{\rm H_{2}}$--$\Sigma_{\rm SFR}$--$\Sigma_{*}$) and dust scaling relations. While the relations with all sub-galactic regions of the galaxies are reasonably tight ($\sigma \lesssim 0.3$ dex), we find that most of the scaling relations exhibit galaxy-to-galaxy variations in normalization and shape. Only two relations of $\Sigma_{\rm dust}$--$\Sigma_{\rm gas}$ and $\Sigma_{\rm dust}$--$\Sigma_{\rm H_{2}}$ do not show noticeable galaxy-to-galaxy variations among our sample galaxies. We further investigate correlations among the scaling relations. We find significant correlations among the normalization of the $\Sigma_{\rm H_{2}}$--$\Sigma_{\rm SFR}$--$\Sigma_{*}$ relations, which suggest that galaxies with higher levels of resolved $\text{H}_{2}$ fraction ($f_{\rm H_{2}}$) tend to have higher levels of resolved star formation efficiency (SFE) and specific star formation rate (sSFR). We also observe that galaxies with higher levels of resolved dust-to-stellar mass ratios tend to have higher levels of resolved sSFR, SFE, and $f_{\rm H_{2}}$. Moreover, we find that galaxies with higher global sSFR and less compact morphology tend to have higher levels of the resolved sSFR, SFE, and $f_{\rm H_{2}}$, which can explain the variations in the normalization of the $\Sigma_{\rm H_{2}}$--$\Sigma_{\rm SFR}$--$\Sigma_{*}$ relationships. Overall, we observe indications of the contributions of both global and local factors in governing the star formation process in galaxies., Comment: 28 pages, 13 figures, accepted for publication in ApJ. piXedfit v1.0 is publicly available at https://github.com/aabdurrouf/piXedfit. Documentation is available at https://pixedfit.readthedocs.io/en/latest/. Some tutorials of practical usages in jupyter notebooks can be found at https://github.com/aabdurrouf/piXedfit/tree/main/examples

The Subaru Hyper Suprime-Cam (HSC) Strategic Survey is the latest-generation multi-band optical imaging survey for galaxy evolution and structure formation. The "Ultra-Deep" component of the HSC survey provides $grizy$ broad-band images over $\sim3.4$ deg$^2$ to detection limits of $\sim26$-28 AB, along with narrow-band images, in the COSMOS and the SXDS fields. These images provide an unprecedented combination of depths and area coverage, for the studies galaxies up to $z\sim7$. However, the lack of coverage at $<4000$ Ang implies incomplete sampling of the rest-frame UV at $z\lesssim 3$, which is critically needed for understanding the buildup of stellar mass in the later cosmic time. We conducted a multi-year CFHT $u^\ast$-band imaging campaign in the two HSC Ultra-Deep fields with CFHT MegaCam. By including shallower archival data, we reach 5-$\sigma$ depths of $u^\ast=28.1$ and 28.4 (AB) at the centers of the COSMOS and SXDS fields, respectively, and $u^\ast=27.7$ and 27.8 in the central 1 deg$^2$ fields. The image quality is $\gtrsim0.90$ arcsec, fairly good for the $u^\ast$ band. Both the photometric and astrometric quality of our data are excellent. We show that the combination of our $u^\ast$-band and HSC data can lead to high-quality photometric redshifts at $z=0$-3, and robust measurements of rest-frame UV on galaxies at $0.4

This paper introduces Miutsu, National Taiwan University's Alexa Prize TaskBot, which is designed to assist users in completing tasks requiring multiple steps and decisions in two different domains -- home improvement and cooking. We overview our system design and architectural goals, and detail the proposed core elements, including question answering, task retrieval, social chatting, and various conversational modules. A dialogue flow is proposed to provide a robust and engaging conversation when handling complex tasks. We discuss the faced challenges during the competition and potential future work.

We propose to adopt statistical regression as the projection operator to enable data-driven learning of the operators in the Mori--Zwanzig formalism. We present a principled method to extract the Markov and memory operators for any regression models. We show that the choice of linear regression results in a recently proposed data-driven learning algorithm based on Mori's projection operator, which is a higher-order approximate Koopman learning method. We show that more expressive nonlinear regression models naturally fill in the gap between the highly idealized and computationally efficient Mori's projection operator and the most optimal yet computationally infeasible Zwanzig's projection operator. We performed numerical experiments and extracted the operators for an array of regression-based projections, including linear, polynomial, spline, and neural-network-based regressions, showing a progressive improvement as the complexity of the regression model increased. Our proposition provides a general framework to extract memory-dependent corrections and can be readily applied to an array of data-driven learning methods for stationary dynamical systems in the literature., Comment: 41 pages, 12 figures; major revision of V2

Dialogue systems are usually categorized into two types, open-domain and task-oriented. The first one focuses on chatting with users and making them engage in the conversations, where selecting a proper topic to fit the dialogue context is essential for a successful dialogue. The other one focuses on a specific task instead of casual talks, e.g., finding a movie on Friday night, or playing a song. These two directions have been studied separately due to their different purposes. However, how smoothly transitioning from social chatting to task-oriented dialogues is important for triggering business opportunities, and there is no public data focusing on such scenarios. Hence, this paper focuses on investigating the conversations starting from open-domain social chatting and then gradually transitioning to task-oriented purposes, and releases a large-scale dataset with detailed annotations for encouraging this research direction. To achieve this goal, this paper proposes a framework to automatically generate many dialogues without human involvement, in which any powerful open-domain dialogue generation model can be easily leveraged. The human evaluation shows that our generated dialogue data has a natural flow at a reasonable quality, showing that our released data has a great potential of guiding future research directions and commercial activities. Furthermore, the released models allow researchers to automatically generate unlimited dialogues in the target scenarios, which can greatly benefit semi-supervised and unsupervised approaches., Comment: ACL 2022

Gene expression is a biochemical process, where stochastic binding and un-binding events naturally generate fluctuations and cell-to-cell variability in gene dynamics. These fluctuations typically have destructive consequences for proper biological dynamics and function (e.g., loss of timing and synchrony in biological oscillators). Here, we show that gene expression noise counter-intuitively accelerates the evolution of a biological oscillator and, thus, can impart a benefit to living organisms. We used computer simulations to evolve two mechanistic models of a biological oscillator at different levels of gene expression noise. We first show that gene expression noise induces oscillatory-like dynamics in regions of parameter space that cannot oscillate in the absence of noise. We then demonstrate that these noise-induced oscillations generate a fitness landscape whose gradient robustly and quickly guides evolution by mutation towards robust and self-sustaining oscillation. These results suggest that noise can help dynamical systems evolve or learn new behavior by revealing cryptic dynamic phenotypes outside the bifurcation point., Comment: 36 pages, 9 figures

A number of recent studies have proposed that linear representations are appropriate for solving nonlinear dynamical systems with quantum computers, which fundamentally act linearly on a wave function in a Hilbert space. Linear representations, such as the Koopman representation and Koopman von Neumann mechanics, have regained attention from the dynamical-systems research community. Here, we aim to present a unified theoretical framework, currently missing in the literature, with which one can compare and relate existing methods, their conceptual basis, and their representations. We also aim to show that, despite the fact that quantum simulation of nonlinear classical systems may be possible with such linear representations, a necessary projection into a feasible finite-dimensional space will in practice eventually induce numerical artifacts which can be hard to eliminate or even control. As a result, a practical, reliable and accurate way to use quantum computation for solving general nonlinear dynamical systems is still an open problem., Comment: 27 pages, 16 figures