arXiv subjects

Nam Phuong Tran

Publications and source records attributed to Nam Phuong Tran.

7 recordsLinked to original sources

arXiv2026

Last-Iterate Guarantees for Online Reinforcement Learning in Structured Constrained MDPs

In safety-critical applications, deployment uses a single policy, whose performance and constraint satisfaction should hold directly rather than only for an average or mixture of training policies. This motivates last-iterate guarantees in constrained reinforcement learning. Recent progress has established such guarantees in exact-gradient or tabular online settings, yet scalable results for structured large-state problems remain open. We develop a general, statistically efficient framework for last-iterate convergence in structured Constrained MDPs (CMDPs). Our analysis separates contraction of the regularised primal-dual dynamics from actor approximation and statistical errors in policy evaluation under online exploration. This enables model-free on- and off-policy learning with structured function approximation: optimistic policy evaluation avoids explicit transition-model construction, while a compact parametric actor avoids maintaining mixtures or histories of past policies. We instantiate the framework for linear CMDPs and general function approximation, obtaining representation-dependent complexity and improved target-accuracy dependence over prior optimistic regularised primal-dual analyses. We further validate the stabilising effect predicted by our theory on a synthetic linear CMDP: the regularised method exhibits stable last-iterate behaviour, whereas its unregularised counterpart shows larger oscillations.

cs.LG↗

arXiv2026

Near-optimal Reconfigurable Intelligent Surface Configuration: Blind Beamforming with Sensing

Blind beamforming has emerged as a promising approach to configure reconfigurable intelligent surfaces (RISs) without relying on channel state information (CSI) or geometric models, making it directly compatible with commodity hardware. In this paper, we propose a new blind beamforming algorithm, so-called Blind Optimal RIS Beamforming with Sensing (\textsc{BORN}), that operates using only received signal strength (RSS). In contrast to existing methods that rely on majority-voting mechanisms, \textsc{BORN} exploits the intrinsic quadratic structure of the received signal-to-noise ratio (SNR). The algorithm proceeds in two stages: \emph{sensing}, where a quadratic model is estimated from RSS measurements, and \emph{optimization}, where the RIS configuration is obtained using the estimated quadratic model. Our novelties are twofold. Firstly, we show for the first time, that \textsc{BORN} can achieve provable near-optimal performance using only $O(N \log_2(N))$ samples, where $N$ is the number of RIS elements. As a by-product of our analysis, we show that quadratic models are learnable under Rademacher feature distributions when the second-order coefficient matrix is low-rank. This result, to our knowledge, has not been established in prior matrix sensing literature. Secondly, extensive simulations and real-world field tests demonstrate that \textsc{BORN} achieves near-optimal performance, substantially outperforming state-of-the-art blind beamforming algorithms, particularly in scenarios with a weak background channel such as non-line-of-sight (NLOS).

eess.SP↗

arXiv2026

Meta-Learning Preferences for Multilingual LLM Alignment

Unequal availability of human preference data across languages poses a significant challenge for aligning large language models in multilingual settings. To address the lack of sufficient data in low-resource language alignment, we propose a meta-learning framework for Reinforcement Learning from Human Feedback and Direct Preference Optimization. By leveraging preference data from other languages, our framework learns a transferable initialization that enables effective adaptation to a target language with minimal data. We provide theoretical guarantees for both the meta-reward modeling and meta-policy optimization settings, and empirically demonstrate the effectiveness of our approach on multilingual benchmarks. In an extremely low-resource setting with only 100 target-language preference samples, our approach achieves up to $28\%$ win-rate improvements over baseline methods, and consistently outperforms baselines across multiple target languages and model scales. Our approaches retain these advantages across different combinations of meta-training languages and varying linguistic distances from the target languages.

cs.CL↗

arXiv2026

Sparse Offline Reinforcement Learning with Corruption Robustness

We investigate robustness to strong data corruption in offline sparse reinforcement learning (RL). In our setting, an adversary may arbitrarily perturb a fraction of the collected trajectories from a high-dimensional but sparse Markov decision process, and our goal is to estimate a near optimal policy. The main challenge is that, in the high-dimensional regime where the number of samples $N$ is smaller than the feature dimension $d$, exploiting sparsity is essential for obtaining non-vacuous guarantees but has not been systematically studied in offline RL. We analyse the problem under uniform coverage and sparse single-concentrability assumptions. While Least Square Value Iteration (LSVI), a standard approach for robust offline RL, performs well under uniform coverage, we show that integrating sparsity into LSVI is unnatural, and its analysis may break down due to overly pessimistic bonuses. To overcome this, we propose actor-critic methods with sparse robust estimator oracles, which avoid the use of pointwise pessimistic bonuses and provide the first non-vacuous guarantees for sparse offline RL under single-policy concentrability coverage. Moreover, we extend our results to the contaminated setting and show that our algorithm remains robust under strong contamination. Our results provide the first non-vacuous guarantees in high-dimensional sparse MDPs with single-policy concentrability coverage and corruption, showing that learning a near-optimal policy remains possible in regimes where traditional robust offline RL techniques may fail.

stat.ML↗

arXiv2025

Symmetric Linear Bandits with Hidden Symmetry

High-dimensional linear bandits with low-dimensional structure have received considerable attention in recent studies due to their practical significance. The most common structure in the literature is sparsity. However, it may not be available in practice. Symmetry, where the reward is invariant under certain groups of transformations on the set of arms, is another important inductive bias in the high-dimensional case that covers many standard structures, including sparsity. In this work, we study high-dimensional symmetric linear bandits where the symmetry is hidden from the learner, and the correct symmetry needs to be learned in an online setting. We examine the structure of a collection of hidden symmetry and provide a method based on model selection within the collection of low-dimensional subspaces. Our algorithm achieves a regret bound of $ O(d_0^{2/3} T^{2/3} \log(d))$, where $d$ is the ambient dimension which is potentially very large, and $d_0$ is the dimension of the true low-dimensional subspace such that $d_0 \ll d$. With an extra assumption on well-separated models, we can further improve the regret to $ O(d_0\sqrt{T\log(d)} )$.

stat.ML↗

arXiv2024

Learning the Expected Core of Strictly Convex Stochastic Cooperative Games

Reward allocation, also known as the credit assignment problem, has been an important topic in economics, engineering, and machine learning. An important concept in reward allocation is the core, which is the set of stable allocations where no agent has the motivation to deviate from the grand coalition. In previous works, computing the core requires either knowledge of the reward function in deterministic games or the reward distribution in stochastic games. However, this is unrealistic, as the reward function or distribution is often only partially known and may be subject to uncertainty. In this paper, we consider the core learning problem in stochastic cooperative games, where the reward distribution is unknown. Our goal is to learn the expected core, that is, the set of allocations that are stable in expectation, given an oracle that returns a stochastic reward for an enquired coalition each round. Within the class of strictly convex games, we present an algorithm named \texttt{Common-Points-Picking} that returns a point in the expected core given a polynomial number of samples, with high probability. To analyse the algorithm, we develop a new extension of the separation hyperplane theorem for multiple convex sets.

cs.GT↗

arXiv2023

Invariant Lipschitz Bandits: A Side Observation Approach

Symmetry arises in many optimization and decision-making problems, and has attracted considerable attention from the optimization community: By utilizing the existence of such symmetries, the process of searching for optimal solutions can be improved significantly. Despite its success in (offline) optimization, the utilization of symmetries has not been well examined within the online optimization settings, especially in the bandit literature. As such, in this paper we study the invariant Lipschitz bandit setting, a subclass of the Lipschitz bandits where the reward function and the set of arms are preserved under a group of transformations. We introduce an algorithm named \texttt{UniformMesh-N}, which naturally integrates side observations using group orbits into the \texttt{UniformMesh} algorithm (\cite{Kleinberg2005_UniformMesh}), which uniformly discretizes the set of arms. Using the side-observation approach, we prove an improved regret upper bound, which depends on the cardinality of the group, given that the group is finite. We also prove a matching regret's lower bound for the invariant Lipschitz bandit class (up to logarithmic factors). We hope that our work will ignite further investigation of symmetry in bandit theory and sequential decision-making theory in general.

cs.LG↗