Searcharxiv⌕ Search

arXiv subjects

Nam Nguyen

Publications and source records attributed to Nam Nguyen.

At least 19 recordsLinked to original sources

EEG-X: Toward Device-Agnostic and Noise-Robust Foundation Model for EEG

Foundation models for EEG analysis are still in their infancy, limited by two key challenges: (1) variability across datasets caused by differences in recording devices and electrode configurations, and (2) the low signal-to-noise ratio (SNR) of EEG, where neural activity is often buried under artifacts and non-brain sources. To address these challenges, we present EEG-X, a device-agnostic and noise-robust foundation model for EEG representation learning. EEG-X introduces a location-based channel embedding that enables robust transfer across heterogeneous EEG devices and electrode layouts. To improve robustness against noise, EEG-X employs a novel noise-aware masking-reconstruction strategy that reconstructs artifact-cleaned signals rather than raw noisy EEG and introduces a novel Dictionary-based Convolutional Transformation (DiCT) that improves reconstruction-based pretraining by comparing signals in a structured feature space instead of directly in the raw signal space. Experiments across datasets collected from diverse EEG devices show that EEG-X consistently outperforms state-of-the-art methods across multiple downstream tasks and demonstrates strong cross-domain generalization when pretraining and downstream datasets differ in electrode layouts and recording configurations paving the way towards foundation models. The models and code are available at https://github.com/Emotiv/EEG-X.

cs.LG↗

Simulating closed- and open-system dynamics of Tavis--Cummings lattices on fault-tolerant quantum computers

Tavis--Cummings lattices couple cavities containing several emitters and host polaritonic many-body dynamics whose real-time correlations are costly to compute classically in two dimensions, especially when unequal emitter relaxation or dephasing rates break permutation symmetry. We analyze logical resources for fault-tolerant closed- and open-system simulation of an atom-resolved lattice with $L$ cavities, $N$ emitters per cavity, and photon cutoff $n\_{\max}$, including target observable estimation. Building on ladder-operator block encoding, we construct circuits for photon hopping and local light--matter exchange and combine them to encode the lattice Hamiltonian. By updating source and destination photon occupations in place, the hopping circuit eliminates the auxiliary $L\lceil\log_2(n\_{\max}+1)\rceil$-qubit photon register used by two-register implementations. Across three uniform-coefficient examples with $16$ to $256$ cavities, the controlled walk uses $4.75$--$8.11%$ fewer Toffoli gates than the paired two-register implementation under matched shared preparation, precision, control, and workspace conventions. For closed systems, we implement real-time evolution with generalized quantum signal processing and bound the cost of estimating retarded cavity spectral functions. For open systems with photon loss, emitter relaxation, and dephasing, we use a sampled first-order dilation and bound the runs and timesteps needed to estimate transient photon statistics such as $g^{(2)}(0)$. With state-preparation and measurement costs explicit, these estimates provide a reference for assessing fault-tolerant light--matter lattice simulation costs.

quant-ph↗

Sharp Non-Asymptotic Analysis of the Penalized Challenger in $β$-EB-TCI for Bernoulli Bandits

Top-two algorithms are simple and effective for fixed-confidence best-arm identification, but their sharp non-asymptotic behavior is still not well understood. We study this question for Bernoulli bandits through $β$-EB-TCI, the empirical-best top-two rule of Jourdan et al., whose challenger minimizes a Bernoulli transportation cost plus a logarithmic penalty in its pull count. We first analyze the stabilized regime, in which the empirical leader has become the true best arm and its sampling fraction stays close to $β$. In this regime, every suboptimal arm is sampled linearly often, and, with high probability, the stopping time is at most $T_β^{\star}(μ)\log(1/δ)$ plus stabilization and lower-order concentration terms, with the exact fixed-$β$ leading coefficient. The main remaining difficulty is therefore to control when the regime is reached. For the original rule, we obtain a non-asymptotic high-probability stopping bound on every Bernoulli instance with a unique best arm. Under a finite-mean sufficient-exploration condition, the bound yields the sharp expected sample complexity; combined with the exploration result of Jourdan et al., this covers pairwise-distinct means under their tie-breaking convention. A mild forced-exploration guard, which uses only $O(\sqrt{Kt})$ pulls up to time $t$, gives a self-contained sharp expected bound for any number of arms with a unique best arm. We also identify a limitation of proof strategies that try to handle equal suboptimal means through a single index-comparison argument.

cs.LG↗

Beyond Expected Values: Risk-Sensitive Planning with Distributional Monte-Carlo Tree Search

Monte-Carlo Tree Search (MCTS) typically stores only scalar mean estimates, which makes it poorly suited for tail-risk objectives such as conditional value-at-risk (CVaR). We introduce Categorical Thompson Sampling with Optimism (CATSO) and Particle Thompson Sampling with Optimism (PATSO), two distributional MCTS algorithms that maintain finite-support empirical backup laws at Q-edges. CATSO represents Q-edge laws with categorical atoms and Dirichlet Thompson sampling, while PATSO represents them with adaptive particles. Both algorithms select actions using the CVaR of a Thompson-sampled Q-edge law plus a polynomial optimism bonus, and propagate scalar continuation values through visit-weighted averages of child Q-edge CVaR scores. For the finite-depth nested CVaR planning target, we prove an $O(n^{-1/2})$ bound on the root value error after $n$ rollouts, which upper-bounds the risk-sensitive simple regret. CATSO adds a fixed-grid discretization bias, while capped PATSO adds a tunable Wasserstein compression bias. Experiments on stochastic planning benchmarks show that distributional Q-edges improve lower-tail return and reduce catastrophic outcomes when mean-optimal and risk-sensitive behavior differ.

cs.LG↗

DP-Rec: Towards Dynamic Patching for Efficient Long-Sequence Recommendation

Transformers have redefined sequential recommendation by effectively modeling dynamic user behaviors and long-range dependencies. However, they remain inherently inefficient: standard architectures operate at a fixed rate, allocating comparable computation to every item in a user's history regardless of its information content. This leads to prohibitive computational overhead on long sequences and increased sensitivity to behavioral noise. To address this, practitioners often resort to lossy sequence compression, staged modeling, or truncation. This limits the model's ability to leverage the full context of long histories during inference. Inspired by the recent success of Byte Latent Transformer, we propose DP-Rec, a dynamic latent patching architecture for recommendation. DP-Rec shifts from item-level modeling to patch-level modeling by segmenting interaction sequences using contrastive entropy surprise to identify informative behavioral boundaries. A lightweight patch encoder compresses these temporally contextualized segments into a reduced set of dynamic latent behavior vectors, which are then processed by a larger latent transformer and decoded for next-item prediction. Extensive experiments show that, under constrained computational budgets, DP-Rec scales effectively to long sequences and achieves a superior efficiency-accuracy trade-off over both non-compressed and fixed-size compression baselines.

cs.LG↗

Variance Driven Exploration: A Provable and Efficient Methodology for Pure Exploration in Highly Stochastic Environments

We propose Variance Driven Exploration (VarDE), a principled approach for pure exploration in highly stochastic environments, where the exploration process is dominated by stochastic variance. VarDE is built on a fundamental principle: sampling effort should be allocated to minimize the uncertainty of the final decision. We formalize the uncertainty of the final decision through a smooth decision function and derive allocation rules that explicitly capture how stochastic noise in individual components affects the reliability of the final output. We apply this methodology to three core problems of pure exploration -- Best Arm Identification (BAI), Monte Carlo Tree Search (MCTS), and Best-Policy Identification (BPI) -- with theoretical guarantees on variance decay and simple regret. Empirically, we demonstrate consistent and significant improvements of VarDE over existing methods, with especially strong gains in highly stochastic environments.

cs.LG↗

Prospects for Quantum Computation in Propellant Design: Assessing the Stability of Cyclic Ozone in Nanoscale Confinement

Cyclic ozone additives have the potential to markedly increase the specific impulse of rocket fuel. This would translate to greater efficiency and reduced costs for space lift by granting more payload per rocket. While practical efforts to capture this isomer have been unsuccessful, it is possible that cyclic ozone would be stabilized in confined geometries. The required synthetic methods are nonetheless difficult to design and require theory-driven inputs that lie beyond the scope of classical methods. Quantum computation has the potential to enable these calculations, though the underlying hardware requirements remain unclear for many practical applications. We present an end-to-end analysis of how quantum methods could support efforts to isolate cyclic ozone via fullerene encapsulation. Our discussion extends beyond asymptotic complexity, reporting both logical- and physical-level resource estimates for ground-state energy determination via quantum phase estimation (QPE), computed using multiple independent resource estimation stacks - Azure Quantum Resource Estimator (AzureQRE), PennyLane's resource estimation framework, and MIT Lincoln Lab's pyLIQTR toolkit - to cross-check resource estimates and bracket realistic overheads. Taken collectively, these data delineate a plausible scale for realistic, computationally-aided molecular design efforts using fault-tolerant quantum computation.

quant-ph↗

Symmetry-based perturbation theory for electronic structure calculations

We develop a multi-reference perturbation theory for electronic structure calculations based on symmetries of the Hamiltonian. The reference Hamiltonian in the symmetry-based perturbation theory (SBPT) is chosen such that it possesses more symmetries than the original Hamiltonian, leading to a larger reduction of computational resources in terms of both the number of configurations in the configuration interaction expansion and the number of required qubits in quantum computing applications. We provide approximate, scalable solutions for the second-order correction, as well as an application to selected configuration interaction. We show that SBPT is an extension of other existing multi-reference perturbation theories and that it can give better results for some molecular systems in a robust way.

quant-ph↗

Dissipative continuation for ground-state preparation at chemical transition states

Simulating chemical reactions exhibits a pronounced unevenness in computational difficulty: while equilibrium reactant and product geometries are often tractable, transition-state (TS) geometries frequently display strong multi-reference character that challenges both classical solvers and coherent quantum state-preparation methods. We introduce a dissipative continuation protocol for preparing electronic ground states near TS geometries within a hybrid classical--quantum workflow. In the intended setting, classical electronic-structure methods supply an approximate TS geometry, a computationally motivated continuation path, and a locally compatible active-space representation along that path. Starting from a warm start at a tractable geometry on the same aligned path, the quantum routine transports the state toward the TS using orbital-gauge-aligned Hamiltonians and engineered dissipative cooling primitives that repeatedly contract population into the instantaneous low-energy sector. We prove that, for continuation paths satisfying a Lipschitz smoothness condition and a localized Eigenstate Thermalization Hypothesis (ETH)-motivated downward-drift condition within the relevant energy window, the ground state at the target geometry can be prepared to total energy error $ε_E$ with total ideal cooling-step complexity $\widetilde{O}(C_{\mathrm{DK}}^2 N_o^2 / ε_E)$. Here $C_{\mathrm{DK}}$ quantifies ground-state rotation along the aligned path. The corresponding logical gate count is obtained by multiplying this primitive count by the cost of implementing one dissipative step, which is polynomial in the block-encoding size of the Lindbladian under standard Lindblad-simulation algorithms. This identifies a structured regime in which dissipative continuation provides a conditional route to ground-state preparation at strongly correlated TS geometries.

quant-ph↗

Preservation of Positive-Definiteness by Bernstein Operators on the Circle

We prove that, for every $n\ge1$, the degree-$n$ Bernstein operator on $[0,π]$ preserves positive-definiteness on the circle $S^1$. Equivalently, if a continuous function on $[0,π]$ defines a positive-definite isotropic kernel on $S^1$, then its Bernstein polynomial approximation of any fixed degree does as well. The proof reduces the problem to the nonnegativity of the cosine coefficients of the Bernstein images $Q_{n,m}=B_n[\cos(mx)]$, which we prove using an explicit coefficient formula and a two-regime positivity argument. We also discuss the higher-dimensional sphere analogue and show that the naive affine Bernstein operator fails to preserve the positive-definite cone already on $S^2$.

math.CA↗

Conservation Laws for Modern Neural Architectures

Understanding gradient descent dynamics is key to explaining the success of over-parameterized models, where implicit bias manifests through conservation laws in gradient flow. While such laws are well understood for linear and ReLU networks, they remain largely unexplored for modern architectures. This work develops a unified framework to characterize conservation laws for contemporary models, including feedforward networks with GELU, SiLU, and SwiGLU activations, multihead attention with sinusoidal and rotary positional encodings, and Mixture-of-Experts architectures under diverse gating designs. Our theoretical findings are supported by experiments that validate the predicted invariants.

cs.LG↗

Quadratically Regularized Optimal Transport: Localization Bounds and Affine Case Analysis

Quadratic regularization has emerged as a potential alternative to the popular entropic regularization in computational optimal transport, offering the theoretical advantage of producing sparse couplings through its hinge density structure. Despite recent progress in one-dimensional settings and general upper bounds, fundamental questions about the localization rate of QOT optimizers around the Monge coupling have remained open. In this work, we establish a general lower bound showing that the support of the QOT optimizer cannot concentrate around the Monge graph faster than order $\varepsilon^{\frac{1}{d+2}}$ in the directed Hausdorff distance, matching the conjectured optimal exponent under standard regularity assumptions in \citet{wiesel2025sparsity}. We also show that the QOT value gap controls the mean-squared deviation $\mathbb E_{π_\varepsilon}\|y-T(x)\|^2$ by the scale of $\varepsilon^{\frac{2}{d+2}}$. As a corollary, in the affine Brenier regime, which includes Gaussian-to-Gaussian transport, we derive a sharp pointwise tube bound of order $\varepsilon^{\frac{1}{d+2}}$ by reducing the problem to self-transport and applying recent self-transport sparsity results. Finally, we validate our theoretical bound with a synthetic experiment in high-dimensional settings.

math.OC↗

Rate-Distortion-Classification Representation Theory for Bernoulli Sources

We study task-oriented lossy compression through the lens of rate-distortion-classification (RDC) representations. The source is Bernoulli, the distortion measure is Hamming, and the binary classification variable is coupled to the source via a binary symmetric model. Building on the one-shot common-randomness formulation, we first derive closed-form characterizations of the one-shot RDC and the dual distortion-rate-classification (DRC) tradeoffs. We then use a representation-based viewpoint and characterize the achievable distortion-classification (DC) region induced by a fixed representation by deriving its lower boundary via a linear program. Finally, we study universal encoders that must support a family of DC operating points and derive computable lower and upper bounds on the minimum asymptotic rate required for universality, thereby yielding bounds on the corresponding rate penalty. Numerical examples are provided to illustrate the achievable regions and the resulting universal RDC/DRC curves.

cs.IT↗

Perception-based Image Denoising via Generative Compression

Image denoising aims to remove noise while preserving structural details and perceptual realism, yet distortion-driven methods often produce over-smoothed reconstructions, especially under strong noise and distribution shift. This paper proposes a generative compression framework for perception-based denoising, where restoration is achieved by reconstructing from entropy-coded latent representations that enforce low-complexity structure, while generative decoders recover realistic textures via perceptual measures such as learned perceptual image patch similarity (LPIPS) loss and Wasserstein distance. Two complementary instantiations are introduced: (i) a conditional Wasserstein GAN (WGAN)-based compression denoiser that explicitly controls the rate-distortion-perception (RDP) trade-off, and (ii) a conditional diffusion-based reconstruction strategy that performs iterative denoising guided by compressed latents. We further establish non-asymptotic guarantees for the compression-based maximum-likelihood denoiser under additive Gaussian noise, including bounds on reconstruction error and decoding error probability. Experiments on synthetic and real-noise benchmarks demonstrate consistent perceptual improvements while maintaining competitive distortion performance.

cs.CV↗

Cross-Domain Lossy Compression via Constrained Minimum Entropy Coupling

This paper studies cross-domain lossy compression through the lens of minimum entropy coupling (MEC) with rate and classification constraints. In this setting, an encoder observes samples from a degraded source domain, while the decoder is required to generate outputs following a prescribed target distribution and to preserve information relevant to a downstream classification task. Motivated by logarithmic-loss distortion, we adopt an information-based objective that maximizes the coupling strength between the source and reconstruction, rather than minimizing a sample-wise distortion. Under common randomness, we formulate a rate-constrained MEC problem (MEC-B) and show that the intermediate representation can be removed without loss of optimality, yielding an equivalent deterministic coupling formulation. For Bernoulli sources, closed-form expressions are derived with and without classification constraints. In addition, we implement a neural restoration framework using quantization, entropy modeling, distribution matching, and classification regularization. Experiments on MNIST super-resolution and SVHN denoising show that increasing the available rate improves classification accuracy and yields more informative reconstructions.

cs.IT↗

Quantum-centric simulation of hydrogen abstraction by sample-based quantum diagonalization and entanglement forging

The simulation of electronic systems is an anticipated application for quantum-centric computers, i.e. heterogeneous architectures where classical and quantum processing units operate in concert. An important application is the computation of radical chain reactions, including those responsible for the photodegradation of composite materials used in aerospace engineering. Here, we compute the activation energy and reaction energy for hydrogen abstraction from 2,2-diphenyldipropane, used as a minimal model for a step in a radical chain reaction. Calculations are performed using a superconducting quantum processor of the IBM Heron family and classical computing resources. To this end, we combine a qubit-reduction technique called entanglement forging (EF) with sample-based quantum diagonalization (SQD), a method that projects the Schrödinger equation into a subspace of configurations sampled from a quantum device. In conventional quantum simulations, a qubit represents a spin-orbital. In contrast, EF maps a qubit to a spatial orbital, reducing the required number of qubits by half. We provide a complete derivation and a detailed description of the combined EF and SQD approach, and we assess its accuracy across active spaces of varying sizes upto (39e,39o).

quant-ph↗

Cooptimizing Safety and Performance Using Safety Value-Constrained Model Predictive Control

Autonomous systems are increasingly deployed in real-world environments, where they must achieve high performance while maintaining safety under state and input constraints. Although Model Predictive Control (MPC) provides a principled framework for constrained optimal control, guaranteeing safety beyond its finite planning horizon remains a fundamental challenge. In this work, we augment MPC with a safety value function-based terminal constraint that enforces membership in a control-invariant safe set at the end of each planning horizon. This formulation enables real-time synthesis of trajectories that are both high-performing and provably safe. We show that, under an exact safety value function and a feasible initialization, the proposed MPC scheme is recursively feasible, thereby ensuring persistent safety. In contrast to traditional terminal set constructions that rely on local linearizations or conservative approximations, our approach incorporates a reachability-based safety value function for terminal constraints, yielding less conservative and more expressive safety guarantees. We validate the proposed framework through simulation and hardware experiments on a Flexiv Rizon 10s manipulator. Results demonstrate improved constraint satisfaction and robustness compared to standard state-constrained MPC and reactive safety filtering, while maintaining competitive task performance. The full implementation and experiments are available on the project website.

cs.RO↗

BEDTime: A Unified Benchmark for Automatically Describing Time Series

Recent works propose complex multi-modal models that handle both time series and language, ultimately claiming high performance on complex tasks like time series reasoning and cross-modal question answering. However, they skip foundational evaluations that such complex models should have mastered. So we ask a simple question: \textit{How well can recent models describe structural properties of time series?} To answer this, we propose that successful models should be able to \textit{recognize}, \textit{differentiate}, and \textit{generate} descriptions of univariate time series. We then create \textbf{\benchmark}, a benchmark to assess these novel tasks, that comprises \textbf{five datasets} reformatted across \textbf{three modalities}. In evaluating \textbf{17 state-of-the-art models}, we find that (1) surprisingly, dedicated time series-language models fall short, despite being designed for similar tasks, (2) vision language models are quite capable, (3) language only methods perform worst, despite many lauding their potential, and (4) all approaches are clearly fragile to a range of real world robustness tests, indicating directions for future work. Together, our findings critique prior works' claims and provide avenues for advancing multi-modal time series modeling.

cs.CL↗