arXiv subjects

Bao Nguyen

Publications and source records attributed to Bao Nguyen.

At least 19 recordsLinked to original sources

arXiv2026

ViP-VL: Vietnamese Self-supervised Speech Pretraining Model with Vector-Quantization Learning

We present ViP-VL, an efficient Vietnamese Self-supervised speech Pretraining model leveraging Vector-quantization Learning. To bridge the gap between high-resolution audio and efficient processing, ViP-VL incorporates Acoustic Stacking and Receptive Field Alignment to enable a synchronized 8x subsampling rate within the ChunkFormer architecture, while further enhancing representation robustness through a specialized Mask Selection Strategy during pretraining on the BEST-RQ framework. Pretrained on 17,000 hours of unlabeled Vietnamese speech, our model establishes new state-of-the-art results across four major downstream tasks: Automatic Speech Recognition, Speech Emotion Recognition, Dialect Classification, and Speaker Verification. To facilitate future research and the development of high-performance Vietnamese speech technologies, we publicly release our pretrained weights and implementation at github.com/khanld/chunkformer.

cs.SD↗

arXiv2026

Adaptive Liquidity in Prediction Markets via Online Learning

Prediction markets rely on liquidity to convert trades into informative prices, yet existing mechanisms fix liquidity ex ante. This restriction enforces a static trade-off between price responsiveness and worst-case loss despite inherently nonstationary trading conditions. We propose a fundamentally different approach that treats liquidity selection itself as an online learning problem. Our mechanism mixes a family of cost-function markets via learnable weights, yielding a single adaptive market that preserves no-arbitrage, bounded worst-case loss, expressiveness, and positive upside. We introduce a hybrid structural risk signal, a per-round objective that quantifies the trade-off between price impact and inventory risk, and show that standard online learning algorithms achieve switching-regret guarantees relative to the best sequence of liquidity regimes in hindsight. Simulations demonstrate that the mechanism adaptively shifts liquidity across regimes in response to both order flow and inventory dynamics. Our results establish a principled framework for adaptive liquidity, connecting prediction market design with online learning.

cs.GT↗

arXiv2026

Adaptive Rollout Allocation for Online Reinforcement Learning with Verifiable Rewards

Sampling efficiency is a key bottleneck in reinforcement learning with verifiable rewards. Existing group-based policy optimization methods, such as GRPO, allocate a fixed number of rollouts for all training prompts. This uniform allocation implicitly treats all prompts as equally informative, and could lead to inefficient computational budget usage and impede training progress. We introduce VIP, a Variance-Informed Predictive allocation strategy that allocates a given rollout budget to the prompts in the incumbent batch to minimize the expected gradient variance of the policy update. At each iteration, VIP uses a lightweight Gaussian process model to predict per-prompt success probabilities based on recent rollouts. These probability predictions are translated into variance estimates, which are then fed into a convex optimization problem to determine the optimal rollout allocations under a hard compute budget constraint. Empirical results show that VIP consistently improves sampling efficiency and achieves higher performance than uniform or heuristic allocation strategies in multiple benchmarks.

cs.LG↗

arXiv2025

Reasoning Planning for Language Models

Selecting an appropriate reasoning method for a given query remains a key challenge in language model generation. Existing approaches typically generate multiple candidate responses and use an aggregation strategy to select the output answer, often assuming that more candidate answers yield higher accuracy. We revisit this assumption through a rigorous theoretical analysis, deriving accuracy bounds for standard aggregation methods under fixed generation distributions and candidate sizes. Building on these insights, we introduce EPIC, an Ensemble Planning with Contrastive learning framework to learn a shared representation space that captures both model reasoning abilities and query-method compatibility. EPIC incorporates our probability bounds as a regularizer in a utility-driven optimization that balances accuracy and computational cost. Experiments on diverse mathematical reasoning tasks show that EPIC consistently selects optimal reasoning methods, improving accuracy while reducing computational overhead. Our code can be found at https://github.com/nguyenngocbaocmt02/EPIC.

cs.LG↗

arXiv2025

Modulated deuteron spectra observed with the Magnetic Recoil neutron Spectrometer at the National Ignition Facility

The Magnetic Recoil Spectrometer (MRS) on the National Ignition Facility is used to measure the neutron spectrum from deuterium-tritium fueled inertial confinement fusion implosions via n-d elastic scattering and magnetic dispersion of recoil deuterons. From the MRS-determined neutron spectrum, the yield ($Y_n$), apparent ion temperature ($T_i$) and areal density ($\rho R$) are determined. However, anomalous energy modulations in recoil deuterons have been observed in several high-yield indirect drive experiments ($Y_n\sim10^{16}-10^{18}$). These observations raise concerns about their potential impact on the MRS-inferred performance metrics. Analytic calculations and particle-in-cell simulations are used to examine the possible beam-plasma instabilities, which indicate the two-stream instability as the driving mechanism behind energy modulations. Based on a statistical analysis of synthetic deuteron spectra, the modulations-induced errors are found to be within the errors of the determined $Y_n$, $T_i$ and $\rho R$ values and thus do not have a significant impact on the MRS measurement.

physics.plasm-ph↗

arXiv2025

VoteGCL: Enhancing Graph-based Recommendations with Majority-Voting LLM-Rerank Augmentation

Recommendation systems often suffer from data sparsity caused by limited user-item interactions, which degrade their performance and amplify popularity bias in real-world scenarios. This paper proposes a novel data augmentation framework that leverages Large Language Models (LLMs) and item textual descriptions to enrich interaction data. By few-shot prompting LLMs multiple times to rerank items and aggregating the results via majority voting, we generate high-confidence synthetic user-item interactions, supported by theoretical guarantees based on the concentration of measure. To effectively leverage the augmented data in the context of a graph recommendation system, we integrate it into a graph contrastive learning framework to mitigate distributional shift and alleviate popularity bias. Extensive experiments show that our method improves accuracy and reduces popularity bias, outperforming strong baselines.

cs.IR↗

arXiv2025

Structured Pruning for Diverse Best-of-N Reasoning Optimization

Model pruning in transformer-based language models, traditionally viewed as a means of achieving computational savings, can enhance the model's reasoning capabilities. In this work, we uncover a surprising phenomenon: the selective pruning of certain attention heads leads to improvements in reasoning performance, particularly on challenging tasks. Motivated by this observation, we propose SPRINT, a novel contrastive learning framework that dynamically selects the optimal head and layer to prune during inference. By aligning question embeddings with head embeddings, SPRINT identifies those pruned-head configurations that result in more accurate reasoning. Extensive experiments demonstrate that our method significantly outperforms traditional best-of-$N$ and random head selection strategies on the MATH500 and GSM8K datasets.

cs.CL↗

arXiv2025

Dynamic Context-Aware Streaming Pretrained Language Model For Inverse Text Normalization

Inverse Text Normalization (ITN) is crucial for converting spoken Automatic Speech Recognition (ASR) outputs into well-formatted written text, enhancing both readability and usability. Despite its importance, the integration of streaming ITN within streaming ASR remains largely unexplored due to challenges in accuracy, efficiency, and adaptability, particularly in low-resource and limited-context scenarios. In this paper, we introduce a streaming pretrained language model for ITN, leveraging pretrained linguistic representations for improved robustness. To address streaming constraints, we propose Dynamic Context-Aware during training and inference, enabling adaptive chunk size adjustments and the integration of right-context information. Experimental results demonstrate that our method achieves accuracy comparable to non-streaming ITN and surpasses existing streaming ITN models on a Vietnamese dataset, all while maintaining low latency, ensuring seamless integration into ASR systems.

cs.CL↗

arXiv2025

Task-driven Layerwise Additive Activation Intervention

Modern language models (LMs) have significantly advanced generative modeling in natural language processing (NLP). Despite their success, LMs often struggle with adaptation to new contexts in real-time applications. A promising approach to task adaptation is activation intervention, which steers the LMs' generation process by identifying and manipulating the activations. However, existing interventions are highly dependent on heuristic rules or require many prompt inputs to determine effective interventions. This paper proposes a layer-wise additive activation intervention framework that optimizes the intervention process, thus enhancing the sample efficiency. We benchmark our framework on various datasets, demonstrating improvements in the accuracy of pre-trained LMs and competing intervention baselines.

cs.CL↗

arXiv2025

Probe-Free Low-Rank Activation Intervention

Language models (LMs) can produce texts that appear accurate and coherent but contain untruthful or toxic content. Inference-time interventions that edit the hidden activations have shown promising results in steering the LMs towards desirable generations. Existing activation intervention methods often comprise an activation probe to detect undesirable generation, triggering the activation modification to steer subsequent generation. This paper proposes a probe-free intervention method FLORAIN for all attention heads in a specific activation layer. It eliminates the need to train classifiers for probing purposes. The intervention function is parametrized by a sample-wise nonlinear low-rank mapping, which is trained by minimizing the distance between the modified activations and their projection onto the manifold of desirable content. Under specific constructions of the manifold and projection distance, we show that the intervention strategy can be computed efficiently by solving a smooth optimization problem. The empirical results, benchmarked on multiple base models, demonstrate that FLORAIN consistently outperforms several baseline methods in enhancing model truthfulness and quality across generation and multiple-choice tasks.

cs.LG↗

arXiv2025

Distributional Surgery for Language Model Activations

Language models, while capable of generating remarkably coherent and seemingly accurate text, can occasionally produce undesirable content, including harmful or toxic outputs. In this paper, we present a new two-stage approach to detect and mitigate undesirable content generations by rectifying activations. First, we train an ensemble of layerwise classifiers to detect undesirable content using activations by minimizing a smooth surrogate of the risk-aware score. Then, for detected undesirable contents, we propose layerwise distributional steering policies that transform the attention heads. These policies are computed through principled semidefinite programming, which aims to minimally perturb the attention distribution while probabilistically guaranteeing the effectiveness of the editions. Empirical evaluations across multiple language models and datasets show that our method outperforms baselines in reducing the generation of undesirable output.

cs.LG↗

arXiv2024

Multi-agent reinforcement learning strategy to maximize the lifetime of Wireless Rechargeable

The thesis proposes a generalized charging framework for multiple mobile chargers to maximize the network lifetime and ensure target coverage and connectivity in large scale WRSNs. Moreover, a multi-point charging model is leveraged to enhance charging efficiency, where the MC can charge multiple sensors simultaneously at each charging location. The thesis proposes an effective Decentralized Partially Observable Semi-Markov Decision Process (Dec POSMDP) model that promotes Mobile Chargers (MCs) cooperation and detects optimal charging locations based on realtime network information. Furthermore, the proposal allows reinforcement algorithms to be applied to different networks without requiring extensive retraining. To solve the Dec POSMDP model, the thesis proposes an Asynchronous Multi Agent Reinforcement Learning algorithm (AMAPPO) based on the Proximal Policy Optimization algorithm (PPO).

cs.LG↗

arXiv2024

Generative Conditional Distributions by Neural (Entropic) Optimal Transport

Learning conditional distributions is challenging because the desired outcome is not a single distribution but multiple distributions that correspond to multiple instances of the covariates. We introduce a novel neural entropic optimal transport method designed to effectively learn generative models of conditional distributions, particularly in scenarios characterized by limited sample sizes. Our method relies on the minimax training of two neural networks: a generative network parametrizing the inverse cumulative distribution functions of the conditional distributions and another network parametrizing the conditional Kantorovich potential. To prevent overfitting, we regularize the objective function by penalizing the Lipschitz constant of the network output. Our experiments on real-world datasets show the effectiveness of our algorithm compared to state-of-the-art conditional distribution learning techniques. Our implementation can be found at https://github.com/nguyenngocbaocmt02/GENTLE.

cs.LG↗

arXiv2024

Sheaf HyperNetworks for Personalized Federated Learning

Graph hypernetworks (GHNs), constructed by combining graph neural networks (GNNs) with hypernetworks (HNs), leverage relational data across various domains such as neural architecture search, molecular property prediction and federated learning. Despite GNNs and HNs being individually successful, we show that GHNs present problems compromising their performance, such as over-smoothing and heterophily. Moreover, we cannot apply GHNs directly to personalized federated learning (PFL) scenarios, where a priori client relation graph may be absent, private, or inaccessible. To mitigate these limitations in the context of PFL, we propose a novel class of HNs, sheaf hypernetworks (SHNs), which combine cellular sheaf theory with HNs to improve parameter sharing for PFL. We thoroughly evaluate SHNs across diverse PFL tasks, including multi-class classification, traffic and weather forecasting. Additionally, we provide a methodology for constructing client relation graphs in scenarios where such graphs are unavailable. We show that SHNs consistently outperform existing PFL solutions in complex non-IID scenarios. While the baselines' performance fluctuates depending on the task, SHNs show improvements of up to 2.7% in accuracy and 5.3% in lower mean squared error over the best-performing baseline.

cs.LG↗

arXiv2024

Comparison of WarpX and GUINEA-PIG for electron positron collisions

As part of the Snowmass'21 planning exercise, the Advanced Accelerator Concepts community proposed developing multi-TeV linear colliders and considered beam-beam effects for these machines. Such colliders operate under a high disruption regime with an enormous number of electron-positron pairs produced from QED effects. Thus, it requires a self-consistent treatment of the fields produced by the pairs, which is not implemented in state-of-the-art beam-beam codes such as GUINEA-PIG. WarpX is a parallel, open-source, and portable particle-in-cell code with an active developer community that models QED processes with photon and pair generation in relativistic laser-beam interactions. However, its application to beam-beam collisions has yet to be fully explored. In this work, we benchmark the luminosity spectra, photon spectra, and coherent production process from WarpX against GUINEA-PIG in the ILC and ultra-tight collision scenarios. Our performance comparison demonstrates a significant speed-up advantage of WarpX, ensuring a more robust and efficient modeling of electron-positron collisions at multi-TeV energies.

physics.acc-ph↗

arXiv2024

Cost-Adaptive Recourse Recommendation by Adaptive Preference Elicitation

Algorithmic recourse recommends a cost-efficient action to a subject to reverse an unfavorable machine learning classification decision. Most existing methods in the literature generate recourse under the assumption of complete knowledge about the cost function. In real-world practice, subjects could have distinct preferences, leading to incomplete information about the underlying cost function of the subject. This paper proposes a two-step approach integrating preference learning into the recourse generation problem. In the first step, we design a question-answering framework to refine the confidence set of the Mahalanobis matrix cost of the subject sequentially. Then, we generate recourse by utilizing two methods: gradient-based and graph-based cost-adaptive recourse that ensures validity while considering the whole confidence set of the cost matrix. The numerical evaluation demonstrates the benefits of our approach over state-of-the-art baselines in delivering cost-efficient recourse recommendations.

cs.LG↗

arXiv2023

Bellman Optimal Stepsize Straightening of Flow-Matching Models

Flow matching is a powerful framework for generating high-quality samples in various applications, especially image synthesis. However, the intensive computational demands of these models, especially during the finetuning process and sampling processes, pose significant challenges for low-resource scenarios. This paper introduces Bellman Optimal Stepsize Straightening (BOSS) technique for distilling flow-matching generative models: it aims specifically for a few-step efficient image sampling while adhering to a computational budget constraint. First, this technique involves a dynamic programming algorithm that optimizes the stepsizes of the pretrained network. Then, it refines the velocity network to match the optimal step sizes, aiming to straighten the generation paths. Extensive experimental evaluations across image generation tasks demonstrate the efficacy of BOSS in terms of both resource utilization and image quality. Our results reveal that BOSS achieves substantial gains in efficiency while maintaining competitive sample quality, effectively bridging the gap between low-resource constraints and the demanding requirements of flow-matching generative models. Our paper also fortifies the responsible development of artificial intelligence, offering a more sustainable generative model that reduces computational costs and environmental footprints. Our code can be found at https://github.com/nguyenngocbaocmt02/BOSS.

cs.CV↗

arXiv2023

AI-assisted Learning for Electronic Engineering Courses in High Education

This study evaluates the efficacy of ChatGPT as an AI teaching and learning support tool in an integrated circuit systems course at a higher education institution in an Asian country. Various question types were completed, and ChatGPT responses were assessed to gain valuable insights for further investigation. The objective is to assess ChatGPT's ability to provide insights, personalized support, and interactive learning experiences in engineering education. The study includes the evaluation and reflection of different stakeholders: students, lecturers, and engineers. The findings of this study shed light on the benefits and limitations of ChatGPT as an AI tool, paving the way for innovative learning approaches in technical disciplines. Furthermore, the study contributes to our understanding of how digital transformation is likely to unfold in the education sector.

cs.CY↗