arXiv subjects

Taehyun Kim

Publications and source records attributed to Taehyun Kim.

At least 37 records · Page 2Linked to original sources

arXiv2026

Ion-Trap Chip Architecture Optimized for Implementation of Quantum Error-Correcting Code

We propose a scalable trapped-ion quantum-computing architecture that efficiently incorporates quantum error correction. The chip design exploits orthogonal qubit connectivity by assigning horizontal trap regions to transversal logical gates and vertical regions to nontransversal gates and syndrome extraction, thereby enabling universal gate operations with minimal ion shuttling and reduced hardware complexity. Using a dedicated software tool, we evaluate the architecture on several benchmark algorithms and scheduling policies for two-dimensional color code of varying code distance. Our results demonstrate that increasing the code distance by two reduces the effective logical two-qubit gate error probability by approximately two orders of magnitude, reaching values as low as $10^{-8}$ with the $[[31, 1, 7]]$ color code. This improvement substantially expands the range of algorithms that can be executed reliably, up to scales of a few thousand logical qubits, depending on the algorithmic structure. Overall, these findings validate the practicality and scalability of the proposed architecture and its control strategies, highlighting a viable route toward fault-tolerant, trapped-ion quantum computing.

quant-ph↗

arXiv2026

A scalable quantum-neural hybrid variational algorithm for ground state estimation

We propose the unitary variational quantum-neural hybrid eigensolver (U-VQNHE), which improves upon the original VQNHE by enforcing unitary neural transformations. The non-unitary nature of VQNHE causes normalization issues and divergence of the loss function during training, leading to exponential scaling of measurement overhead with qubit number. U-VQNHE resolves these issues, significantly reduces required measurements, and retains improved accuracy and stability over standard variational quantum eigensolvers.

quant-ph↗

arXiv2026

Noise-adaptive hybrid quantum convolutional neural networks based on depth-stratified feature extraction

Hierarchical quantum classifiers, such as quantum convolutional neural networks (QCNNs), represent recent progress toward designing effective and feasible architectures for quantum classification. However, their performance on near-term quantum hardware remains highly sensitive to noise accumulation across circuit depth, calling for strategies beyond circuit-architecture design alone. We propose a noise-adaptive hybrid QCNN that improves classification under noise by exploiting depth-stratified intermediate measurements. Instead of discarding qubits removed during pooling operations, we measure them and use the resulting outcomes as classical features that are jointly processed by a classical neural network. This hybrid hierarchical design enables noise-adaptive inference by integrating quantum intermediate measurements with classical post-processing. Systematic experiments across multiple circuit sizes and noise settings, including hardware-calibrated noise models derived from IBM Quantum backend data, demonstrate more stable convergence, reduced loss variability, and consistently higher classification accuracy compared with standard QCNNs. Moreover, we observe that this performance advantage significantly amplifies as the circuit size increases, confirming that the hybrid architecture mitigates the scaling limitations of standard architectures. Notably, the multi-basis measurement variant attains performance close to the noiseless limit even under realistic noise. While demonstrated for QCNNs, the proposed depth-stratified feature extraction applies more broadly to hierarchical quantum classifiers that progressively discard qubits.

quant-ph↗

arXiv2026

Phase-Space Topology in a Single-Atom Synthetic Dimension

We investigate topological features in the synthetic Fock-state lattice (FSL) of a single-atom system described by the quantum Rabi model. By diagonalizing the Hamiltonian, we identify a zero-energy defect state localized at a domain wall of the FSL, whose spin polarization is topologically protected. To address the challenge of applying band topology to the FSL, we introduce a physically motivated and directly measurable topological invariant based on phase-space geometry-the phase-space winding number. We show that the Zak phase, computed using a phase-space parameter, is related to the invariant. This quantized geometric phase reflects the spin polarization of the defect state, demonstrating a bulk-boundary correspondence. The resulting phase-space topology reveals the emergence of single-atom dressed states with contrasting properties-topologically protected spin states and driving-tunable bosonic states. Our results establish phase-space topology as a novel framework for exploring topological physics in single-atom synthetic dimensions, uncovering quantum-unique topological protection distinct from classical analogs.

quant-ph↗

arXiv2025

The impact of bars on the properties of HII regions in the TIMER survey

In this study we perform a comparative analysis of the properties of the HII regions located in different areas of barred galaxies, with the aim of investigating the impact of bars on the physical properties of the ionised gas. Based on integral field spectroscopy data for 17 barred galaxies covering approximately the central 6x6 kpc, we detect a total of 2200 HII regions, of which 331 are located within the nuclear disc (also known as circumnuclear regions), 661 in the bar region, and 1208 in the disc. Among the physical properties of the HII regions, we explore the O/H and N/O abundances, H$α$ luminosity, dust extinction, electron density, and H$α$ equivalent width. We find clear differences in the properties of the HII regions between the nuclear disc, the bar and the disc, that could be explained by an enhancement in the molecular gas concentration in the central parts driven by bar-induced gas flows. As this gas is channelled towards the galaxy centre, the most extreme values in the analysed properties are found for the circumnuclear HII regions. Unlike the bar strength, galaxy mass does seem to affect the properties of the HII regions, with massive galaxies presenting higher values in most of the properties, possibly due to the increased amount of gas in these systems. This study provides evidence that the bar-driven redistribution of material within the galaxy inner parts causes significant differences in the HII region properties depending on their location within the galaxies.

astro-ph.GA↗

arXiv2025

Bar ages derived for the first time in nearby galaxies: Insights on secular evolution from the TIMER sample

Once galaxies settle their discs and become self-gravitating, stellar bars can form, driving the subsequent evolution of their host galaxy. Determining the ages of bars can therefore shed light on the epoch of the onset of secular evolution. In this work, we apply the first broadly applicable methodology to derive bar ages to a sample of 20 nearby galaxies. The method is based on the co-eval build-up of nuclear structures and bars and involves using IFS data from the MUSE instrument on VLT to disentangle the SFH of the nuclear disc from the background population. This allows us to derive the formation epoch of the nuclear disc and, thus, of the bar. We estimate the bar formation epoch of nearby galaxies - mostly from the TIMER survey-, creating the largest sample of galaxies with known bar ages to date. We find bar formation epochs between 1 and 13 Gyr ago, illustrating how disc-settling and bar formation are processes that first took place in the early Universe and are still taking place in some galaxies. We infer the bar fraction over cosmological time with our sample, finding remarkable agreement with that obtained from direct studies of galaxies at high redshifts. Additionally, for the first time, we can investigate secular evolution processes taking into account the ages of bars. Our results agree with the scenario in which bars aid the quenching of the host galaxy, with galaxies hosting older bars tending to be more "quenched". We also find that older bars tend to be longer, stronger, and host larger nuclear discs. Furthermore, we find evidence of the nuclear disc stellar mass build-up over time. On the other hand, we find no evidence of downsizing playing a role in bar formation, since we find that bar age is independent of galaxy stellar mass. With the means to estimate bar ages, we can begin to understand better when and how bars shape the observed properties of disc galaxies.

astro-ph.GA↗

arXiv2025

Motif-2-12.7B-Reasoning: A Practitioner's Guide to RL Training Recipes

We introduce Motif-2-12.7B-Reasoning, a 12.7B parameter language model designed to bridge the gap between open-weight systems and proprietary frontier models in complex reasoning and long-context understanding. Addressing the common challenges of model collapse and training instability in reasoning adaptation, we propose a comprehensive, reproducible training recipe spanning system, data, and algorithmic optimizations. Our approach combines memory-efficient infrastructure for 64K-token contexts using hybrid parallelism and kernel-level optimizations with a two-stage Supervised Fine-Tuning (SFT) curriculum that mitigates distribution mismatch through verified, aligned synthetic data. Furthermore, we detail a robust Reinforcement Learning Fine-Tuning (RLFT) pipeline that stabilizes training via difficulty-aware data filtering and mixed-policy trajectory reuse. Empirical results demonstrate that Motif-2-12.7B-Reasoning achieves performance comparable to models with significantly larger parameter counts across mathematics, coding, and agentic benchmarks, offering the community a competitive open model and a practical blueprint for scaling reasoning capabilities under realistic compute constraints.

cs.AI↗

arXiv2025

Two-Mode Bosonic State Tomography with Single-Shot Joint-Parity Measurement of a Trapped Ion

The full characterization of a continuous-variable quantum system is a challenging problem. For the trapped-ion system, a number of methods of measuring the quantum states have been developed, including the measurement of the Q quasiprobability function and the density-matrix elements in the Fock basis, but these approaches are often slow and difficult to scale to multimode states. Here, we demonstrate a novel and powerful scheme for measuring a continuous-variable quantum state that uses the direct single-shot measurement of the joint parity of the phonon states of a trapped ion. We drive a spin-dependent bichromatic beam-splitter interaction that coherently exchanges phonons between different harmonic oscillator modes of the ion. This interaction encodes the joint-parity information into the relative phase between the two spin states, enabling measurement of the combined phonon-number parity across multiple modes in a single shot. Leveraging this capability, we directly measure multimode Wigner quasiprobability distributions to perform quantum state tomography of an entangled coherent state, and calculate various quantum informational quantities with a model-based estimation of the density matrix. We further show that the single-shot joint-parity measurement can be used to detect parity-flip errors in real time. By postselecting the parity-measurement outcomes, we experimentally demonstrate the partial recovery of coherence, effectively implementing an error-mitigation technique. Lastly, we identify the various sources of error affecting the fidelity of the spin-dependent beam-splitter operation and study the feasibility of high-fidelity operations. The interaction studied in this work can be extended to more than two modes, and is highly relevant to continuous-variable quantum computing and quantum metrology.

quant-ph↗

arXiv2025

The evolution of the bar fraction and bar lengths in the last 12 billion years

We investigate the evolution of the bar fraction and length using an extended JWST NIRCam imaging dataset of galaxies at $1 \leq z \leq 4$. We assess the wavelength dependence of the bar fraction and bar length evolution by selecting a nearly mass-complete CEERS disc sample and performing visual classifications on the short (F200W) and long (F356W+F444W) wavelength channels. A similar bar fraction is observed for both samples, and combined, we find a declining bar fraction from $0.16^{+0.03}_{-0.03}$ to $0.07^{+0.03}_{-0.01}$ over the redshift range. No evolution in the F356W+F444W bar length is measured, with a mean of 3.6 kpc. A slight increase of $\sim 1$ kpc towards $z = 1$ is measured in the F200W sample, with a mean of 2.9 kpc. We find that the correlation between bar length and galaxy mass, for massive galaxies at $z < 1$, is unseen at $z > 1$. By incorporating barred galaxies at $z<1$, we show that there is a modest increase in the bar length ($\approx 2$ kpc) towards $z=0$, but bars longer than $\approx8$ kpc are only found at $z<1$. We show that bars and discs grow in tandem, for the bar length normalised by disc size does not evolve. Not only is a significant population of bars forming beyond $z = 1$, but our results also show that some of these bars are as long and strong as the average bar at $z\approx0$.

astro-ph.GA↗

arXiv2025

Motif 2 12.7B technical report

We introduce Motif-2-12.7B, a new open-weight foundation model that pushes the efficiency frontier of large language models by combining architectural innovation with system-level optimization. Designed for scalable language understanding and robust instruction generalization under constrained compute budgets, Motif-2-12.7B builds upon Motif-2.6B with the integration of Grouped Differential Attention (GDA), which improves representational efficiency by disentangling signal and noise-control attention pathways. The model is pre-trained on 5.5 trillion tokens spanning diverse linguistic, mathematical, scientific, and programming domains using a curriculum-driven data scheduler that gradually changes the data composition ratio. The training system leverages the MuonClip optimizer alongside custom high-performance kernels, including fused PolyNorm activations and the Parallel Muon algorithm, yielding significant throughput and memory efficiency gains in large-scale distributed environments. Post-training employs a three-stage supervised fine-tuning pipeline that successively enhances general instruction adherence, compositional understanding, and linguistic precision. Motif-2-12.7B demonstrates competitive performance across diverse benchmarks, showing that thoughtful architectural scaling and optimized training design can rival the capabilities of much larger models.

cs.CL↗

arXiv2025

Dark gaps and resonances in barred galaxies

Dark gaps, low surface brightness regions along the bar minor axis, are expected to form as a consequence of secular evolution in barred galaxies. Although several studies have proposed links between dark gap locations and dynamical resonances, the results remain inconclusive. Using DESI Legacy Imaging Survey data, we find that approximately 61% of barred galaxies exhibit pronounced dark gaps. We compare the location of dark gaps with resonance radii derived from the Tremaine-Weinberg method applied to MaNGA data for the same galaxies. Our analysis shows that dark gaps do not preferentially form at specific resonances. Instead, their locations correlate with $\mathcal{R}$ $\equiv$ $R_{CR}/R_{Bar}$: slow bars tend to show shorter dark gap radii, while fast bars show longer ones. This trend reflects a tight relation between bar length and dark gap radius. However, when barred galaxies are classified by their ring morphology, certain types exhibit dark gaps that align with specific resonances. Notably, dark gaps located between the inner and outer rings are closely associated with the corotation radius. In galaxies with two dark gaps along the bar minor axis profile, the inner dark gap typically aligns with the ultraharmonic resonance, and the outer dark gap corresponds to the corotation radius. These findings suggest that some morphological types share similar $\mathcal{R}$ values and exhibit dark gaps near specific resonances. Thus, dark gaps may serve as proxies for dynamical resonances only in certain systems. Our findings may help explain the discrepancies observed in earlier studies.

astro-ph.GA↗

arXiv2025

Multi-channel convolutional neural quantum embedding

Classification using variational quantum circuits is a promising frontier in quantum machine learning. Quantum supervised learning (QSL) applied to classical data using variational quantum circuits involves embedding the data into a quantum Hilbert space and optimizing the circuit parameters to train the measurement process. In this context, the efficacy of QSL is inherently influenced by the selection of quantum embedding. In this study, we introduce a classical-quantum hybrid approach for optimizing quantum embedding beyond the limitations of the standard circuit model of quantum computation (i.e., completely positive and trace-preserving maps) for general multi-channel data. We benchmark the performance of various models in our framework using the CIFAR-10 and Tiny ImageNet datasets and provide theoretical analyses that guide model design and optimization.

quant-ph↗

arXiv2025

Expanding Foundational Language Capabilities in Open-Source LLMs through a Korean Case Study

We introduce Llama-3-Motif, a language model consisting of 102 billion parameters, specifically designed to enhance Korean capabilities while retaining strong performance in English. Developed on the Llama 3 architecture, Llama-3-Motif employs advanced training techniques, including LlamaPro and Masked Structure Growth, to effectively scale the model without altering its core Transformer architecture. Using the MoAI platform for efficient training across hyperscale GPU clusters, we optimized Llama-3-Motif using a carefully curated dataset that maintains a balanced ratio of Korean and English data. Llama-3-Motif shows decent performance on Korean-specific benchmarks, outperforming existing models and achieving results comparable to GPT-4.

cs.CL↗

arXiv2025

The Tremaine-Weinberg method at high redshifts

This paper examines the reliability of the Tremaine-Weinberg (TW) method in measuring the pattern speed of barred galaxies at high redshifts. Measuring pattern speeds at high redshift may help to shed light on the time evolution of interactions between galactic bars and dark matter halos. The TW method has been extensively employed for nearby galaxies, and its accuracy in determining bar pattern speeds has been validated through numerical simulations. For nearby galaxies, the method yields acceptable results when the inclination angle of the galaxy and the position angle of the bar fall within appropriate ranges. However, the application of the TW method to high-redshift galaxies remains unexplored in both observations and simulations. For this study we generated mock observations of barred galaxies from the TNG50 cosmological simulation. These simulated observations were tailored to mimic the integral field unit (IFU) spectroscopy data that the Near-Infrared Spectrograph (NIRSpec) on the James Webb Space Telescope (JWST) would capture at a redshift of $z\simeq 1.2$. By applying the TW method to these mock observations and comparing the results with the known pattern speeds, we demonstrate that the TW method performs adequately for barred galaxies whose bars are sufficiently long to be detected by JWST at high redshifts. This work opens a new avenue for applying the TW method to investigate the properties of high-redshift barred galaxies.

astro-ph.GA↗

arXiv2025

Search for Slow Bars in Two Barred Galaxies with Nuclear Structures: NGC 6951 and NGC 7716

We investigate two barred galaxies with nuclear structures, NGC 6951 and NGC 7716, to examine whether they host slow bars. Using Gemini/GMOS long-slit spectroscopy, we calculate the bar pattern speed with the Tremaine-Weinberg method and detect kinematically decoupled nuclear disks in both galaxies. We also measure the bar length and strength using Pan-STARRs images and identify a nuclear ring in NGC 6951 and a nuclear bar in NGC 7716 from HST/PC images. Our results indicate that NGC 6951 hosts a slow, long, and strong bar, which likely evolved through interactions with the dark matter halo and contributed to the formation of both the nuclear disk and ring. We also find hints of a rapidly rotating oval structure within the primary bar, although it is not clearly seen in the imaging data. In contrast, the primary bar in NGC 7716 is too weak to be classified as a barred galaxy, while its nuclear disk and nuclear bar are unusually large, possibly due to tidal interactions or the weakness of the primary bar. These findings suggest that slow bars may be more observed in galaxies with nuclear structures and highlight the often underappreciated role of galaxy interactions in bar evolution.

astro-ph.GA↗

arXiv2025

Optimal Test-Data Piling in HDLSS Classification with Covariance Heterogeneity

This work addresses a longstanding question in high-dimensional linear classification: Is perfect classification achievable in heterogeneous covariance structures? We focus on the phenomenon of data piling, where projected data points collapse onto discrete values. We provide a comprehensive characterization of two distinct types of data piling. The first type of data piling refers to the phenomenon where projecting the training data onto a certain direction yields exactly two distinct values-one for each class. This occurs universally when the data dimension $p$ exceeds the sample size $n$. The second type concerns independent test data and arises asymptotically as $p \to \infty$ with fixed $n$. While previous work established the existence of such double data piling under homogeneously spiked covariance structures using negatively ridged classifiers, our analysis extends to the more general and realistic case of heterogeneous covariance. We identify an optimal direction among all piling directions that maximizes the separation between test data piles, which is called the Second Maximal Data Piling direction. An algorithm based on data splitting is proposed to compute this direction using only training data. Our analysis reveals a key insight: the main obstacle to discovering this direction is the imbalance of the tail eigenvalues, rather than differences in spike count, spike magnitude, or the alignment of leading eigenspaces. Extensive simulations confirm our theoretical results and demonstrate the effectiveness of the proposed classifier across a wide range of high-dimensional scenarios.

math.ST↗

arXiv2025

Dissecting Bar-Induced Stellar Kinematics in Disk Galaxies: The Bisymmetric Model and Rotation Curve Modifications

We analyze bars formed in $N$-body simulations to investigate two key aspects of stellar kinematic structure of barred galaxies: the angular distributions of the radial and azimuthal components of stellar velocities, and the impact of bars on rotation curves. We find that stars on bar-supporting $x_1$-like orbits exhibit characteristic sawtooth-like radial velocity patterns and arch-like tangential velocity patterns as a function of azimuth. In contrast, stars on box and disk orbits show little azimuthal variation, effectively smoothing the overall velocity distribution. When averaged over all orbital families, the resulting kinematics are broadly consistent with the bisymmetric model of Spekkens & Sellwood, with the amplitudes of bar-induced velocity perturbations increasing with bar strength. In addition, bars amplify the radial pressure gradient associated with enhanced random stellar motions, leading to a noticeable reduction in the mean rotational velocity. This effect becomes more pronounced with increasing bar strength, resulting in a shallower rotation curve within the bar region. We discuss our results in the context of the kinematic properties of observed barred galaxies.

astro-ph.GA↗

arXiv2025

Optimising the analysis of emission lines in galaxies: the case of the MUSE TIMER galaxy NGC 613

Galaxy evolution is driven by spatially distributed processes with varying timescales. Integral field spectroscopy provides spatially-resolved information about these processes. Nevertheless, disentangling these processes, which are related to both the underlying stellar populations and the interstellar medium can be challenging. We present a case study on NGC~613, observed with MUSE (Multi-Unit Spectroscopic Explorer) for the TIMER (Time Inference with MUSE in Extragalactic Rings) project, a local barred galaxy, which shows several gas ionisation mechanisms and is rich in both large and inner-scale stellar structures. We develop a set of steps to overcome fundamental problems in the modelling of emission lines with multiple components, together with the characterisation of the stellar populations. That results in the disentanglement of the gas ionisation mechanisms and kinematics, along with an optimal parametrisation for star formation history recovery. Our analysis reveals evidence of gas inflows, which are associated with the bar dust lanes traced with \textit{Hubble} Space Telescope (HST). In addition, we show the gas kinematics in a central biconical outflow, which is aligned with a radio jet observed with Very Large Array (VLA). The emission line provides estimates of electron density, gas-phase metallicity, and the mass outflow rate, allowing us to distinguish intertwined ionisation mechanisms and to identify a part of the multiphase gas cycle in NGC 613. It traces the gas kinematics from the bar lanes to inner scale gas reservoirs, where it can eventually trigger star formation or AGN activity, as observed in the outflow.

astro-ph.GA↗