Searcharxiv⌕ Search

arXiv subjects

Hao Wang

Publications and source records attributed to Hao Wang.

At least 217 records · Page 12Linked to original sources
arXiv2026

Asynchronous Fractional Multi-Agent Deep Reinforcement Learning for Age-Minimal Mobile Edge Computing

In the realm of emerging real-time networked applications such as cyber-physical systems (CPS), the Age of Information (AoI) has emerged as a pivotal metric for evaluating timeliness. To meet the high computational demands, such as those in smart manufacturing within CPS, mobile edge computing (MEC) presents a promising solution for optimizing computing and reducing AoI. In this work, we study the timeliness of compute-intensive updates and explore jointly optimizing the task updating (when to generate a task) and offloading (where to process a task) policies to minimize AoI. Specifically, we consider edge load dynamics and formulate a task scheduling problem to minimize the expected time-average AoI. Solving this problem is challenging due to the fractional objective introduced by AoI and the asynchronous decision-making of the semi-Markov game (SMG). To this end, we propose a fractional reinforcement learning (RL) framework. We begin by introducing a fractional single-agent RL framework and establish its linear convergence rate. Building on this, we develop a fractional multi-agent RL framework, extend Dinkelbach's method, and demonstrate its equivalence to the inexact Newton's method. Furthermore, we provide the conditions under which the framework achieves linear convergence to the Nash equilibrium (NE). To tackle the challenge of asynchronous decision-making in the SMG, we further design an asynchronous model-free fractional multi-agent RL algorithm, where each mobile device can determine the task updating and offloading decisions without knowing the real-time system dynamics and decisions of other devices. Experimental results show that when compared with the best existing baseline algorithm, our proposed algorithm reduces the average AoI by up to 50.6%.

cs.LG↗
arXiv2026

Loong: Synthesize Long Chain-of-Thoughts at Scale through Verifiers

Recent advances in Large Language Models (LLMs) have shown that their reasoning capabilities can be significantly improved through Reinforcement Learning with Verifiable Reward (RLVR), particularly in domains like mathematics and programming, where ground-truth correctness can be automatically evaluated. However, extending this success to other reasoning-intensive domains remains challenging due to the scarcity of high-quality, verifiable datasets and the high cost of human supervision. In this work, we introduce the Loong Project: an open-source framework for scalable synthetic data generation and verification across a diverse range of reasoning-intensive domains. The framework consists of two key components: (1) LoongBench, a curated seed dataset containing 8,729 human-vetted examples across 12 domains (e.g., Advanced Mathematics, Chemistry, Logic), each paired with executable code and rich metadata; and (2) LoongEnv, a modular synthetic data generation environment that supports multiple prompting strategies to produce new question-answer-code triples. Together, these components form an agent-environment loop that enables reinforcement learning, where an LLM-based agent is rewarded for generating Chain-of-Thought (CoT) solutions that align with code-executed answers. Empirically, we benchmark LoongBench on a broad suite of both open-source and proprietary LLMs to evaluate domain coverage and reveal performance bottlenecks. In addition, we conduct a comprehensive analysis of synthetic data generated by LoongEnv, examining correctness, difficulty, and diversity. Code and documentation are available at https://github.com/camel-ai/loong.

cs.LG↗
arXiv2026

Optimising Neural Speech Codecs for 300bps Communication using Reinforcement Learning

In bandwidth-constrained communication such as satellite and underwater channels, speech must often be transmitted at ultra-low bitrates where intelligibility is the primary objective. At such extreme compression levels, codecs trained with acoustic reconstruction losses tend to allocate bits to perceptual detail, leading to substantial degradation in word error rate (WER). This paper proposes ClariCodec, a neural speech codec operating at 300 bits per second (bps) that reformulates quantisation as a stochastic policy, enabling reinforcement learning (RL)-based optimisation of intelligibility. Specifically, the encoder is fine-tuned using WER-driven rewards while the acoustic reconstruction pipeline remains frozen. Even without RL, ClariCodec achieves 4.64% WER on the LibriSpeech test-clean set at 300 bps, already competitive with codecs operating at higher bitrates. Further RL fine-tuning reduces WER to 3.55% on test-clean, corresponding to a 23.5% relative reduction while preserving perceptual quality. In addition, we adapt ClariCodec to a streaming configuration and show that the proposed RL-based optimisation remains effective under streaming constraints, achieving 4.53% WER on test-clean with a theoretical latency of 374 ms.

cs.SD↗
arXiv2026

Cross-Tokenizer On-Policy Distillation via Byte-Prefix Marginalization

Open-weight language models from different families exhibit complementary capabilities, motivating their consolidation into a compact student through on-policy distillation (OPD). However, full-vocabulary OPD typically assumes a shared tokenizer, while existing cross-tokenizer methods may discard teacher probability mass or assign it to student tokens with unrelated content. We introduce Byte-Prefix Marginalization (BPM), which re-expresses the teacher's next-token distribution over the student vocabulary in a shared byte space. Specifically, BPM assigns each teacher token's probability to the longest student token whose byte representation is a prefix of the teacher token's bytes, aggregates mass mapped to the same student token, and places otherwise unmatched mass in an explicit residual category. This produces a vocabulary-complete, byte-aligned, and mass-preserving target for dense OPD. The target exactly recovers the teacher-induced byte-prefix marginal when the relevant prefix does not span multiple teacher tokens (a condition satisfied at more than 99% of training positions) and uses a mass-preserving, chain-factorized lower bound otherwise. Across Qwen3-32B, GLM-Z1-9B-0414, and MiniMax-M2.7 as teachers, BPM consistently outperforms current cross-tokenizer methods on six mathematics and programming benchmarks, improving six-benchmark avg@8 by 3.7-6.6 points over the strongest baselines.

cs.LG↗
arXiv2026

Twins: Learn to Predict Unified Representations with Focal Loss

Unified multimodal models seek a shared visual token space that supports both multimodal understanding and image generation. Discrete methods unify the interface via a shared codebook, whereas continuous pipelines often rely on two disparate representations -- semantic features (e.g., ViT) for understanding and low-level latents (e.g., VAE) for synthesis -- resulting in mismatched latent spaces. We propose Twins, a unified continuous token space formed by channel-wise concatenating ViT and VAE features on the same token grid, so the sequence length is unchanged and attention cost does not increase. However, jointly modeling Twins in a Diffusion Transformer exposes a severe optimization imbalance: the model fits the ViT component well but struggles to match the VAE latent distribution. We trace this imbalance to three sources of heterogeneity: frequency bias, intrinsic dimensionality, and condition-aligned vs condition-independent uncertainty. To address it, we adapt a focal regression objective for flow matching that upweights large-error VAE dimensions, better balancing optimization across the ViT and VAE components. On ImageNet, this yields up to 10.57 gFID gain over naive MSE loss without classifier-free guidance. Twins also performs competitively on multimodal understanding benchmarks and improves reconstruction fidelity, narrowing the gap between understanding- and generation-oriented representations.

cs.CV↗
arXiv2026

Probabilistic Residual Learning for Online Recommendations

Modern recommender systems are typically based on deep learning (DL) models, where a dense encoder learns representations of users and items. As a result, these systems often suffer from the black-box nature and computational complexity of the underlying models, making it difficult to systematically enhance their recommendation capabilities. To address this problem, we propose Probabilistic Residual Learning (PRL), a causal Bayesian recommendation model that models the residual between ground-truth and base predictions, enabling targeted refinement of existing systems. Specifically, PRL (1) probabilistically groups users for localized residual modeling, (2) models domain-level confounders that influence user and item representations, and (3) aggregates cluster-specific residual predictions over the confounders using do-calculus. Experiments demonstrate that our plug-and-play PRL is compatible with various base deep learning recommender systems, improving their performance while automatically discovering meaningful user clusters.

cs.IR↗
arXiv2026

Hierarchically discriminating Haar-randomness in quantum states from a black-box device

The concept of randomness in quantum computing has been central to constructing benchmarking tools, cryptographic protocols, as well as a proof of beyond-classical computation. Discerning whether quantum states (or unitaries) are randomly distributed is a computational task that requires an enormous amount of quantum computational resources. This work addresses such a challenge by introducing a hierarchical discrimination algorithm to efficiently test if a set of states $S$ generated from a black-box quantum device with an unknown distribution is (in)compatible with a random distribution. To this end, we reduce the complexity of the problem by selecting an observable with known spectrum to study the statistical properties of its expectation values with respect to the quantum states from an unknown (black-box) quantum device. Concurrently, we use our first technical result, a connection between Haar-randomness and the Dirichlet distribution, to analytically compute Haar-random moments of the observable. Our Haar-random discriminator test is then simply to compare those statistical moments, such that if $S$ fails the test, it is enough to state that the quantum device does not output randomly distributed states. Else, we can not (yet) confirm that the states follow a Haar-random distribution. We further provide an extension to this algorithm by permutation- and unitary-equivalent randomization of the observable at increasing computational resources, which allows us to more accurately state whether \(S\) is compatible with Haar-randomness. We envision the use of the discriminator test as a quantum device benchmark, by discriminating whether the states generated are incompatible with Haar-randomness.

quant-ph↗
arXiv2026

STEREOFLOW: Progressive Stereo Matching with StereoDiT and Transition Flow Matching

Stereo matching is a fundamental task in 3D reconstruction. Despite remarkable advances, the prevailing paradigms formulate stereo matching as a deterministic regression problem, collapsing the multimodal distribution modeling into a single-point estimation. This formulation suffers from a regression-to-mean bias, frequently struggling with ambiguous regions. In contrast, we introduce a prior-guided generative framework that integrates deterministic matching regression and generative distribution modeling within a complementary formulation. Built upon this formulation, we introduce StereoFlow through three key components: (i) a two-stage progressive cascade matching network that progressively produces multi-resolution stereo conditions with complementary matching cues; (ii) a pixel diffusion transformer (termed StereoDiT) with a frequency-decoupled architecture for modeling correspondence ambiguity; (iii) a few-step flow matching objective (termed Transition Flow Matching) for efficient optimization. In summary, \textsc{\textbf{StereoFlow}} achieves strong geometric consistency and rich fine-grained details in ill-posed, discontinuous regions and under zero-shot generalization. Extensive experiments demonstrate that the proposed StereoFlow establishes multiple state-of-the-art results across benchmarks, including Scene Flow, KITTI, ETH3D, and Middlebury.

cs.CV↗
arXiv2026

FedCod: An Efficient Communication Protocol for Cross-Silo Federated Learning with Coding

Federated Learning (FL) is an innovative distributed machine learning paradigm that enables multiple parties to collaboratively train a model without sharing their raw data, thereby preserving data privacy. Communication efficiency concerns arise in cross-silo FL, particularly due to the network heterogeneity and fluctuations associated with geo-distributed silos. Most existing solutions to these problems focus on algorithmic improvements that alter the FL algorithm but sacrificing the training performance. How to address these problems from a network perspective that is decoupled from the FL algorithm remains an open challenge. In this paper, we propose FedCod, a new application layer communication protocol designed for cross-silo FL. FedCod transparently utilizes a coding mechanism to enhance the efficient use of idle bandwidth through client-to-client communication, and dynamically adjusts coding redundancy to mitigate network bottlenecks and fluctuations, thereby improving the communication efficiency and accelerating the training process. In our real-world experiments, FedCod demonstrates a significant reduction in average communication time by up to 62% compared to the baseline, while maintaining FL training performance and optimizing inter-client communication traffic.

cs.DC↗
arXiv2026

GATE-3D: Geometry-Aware Test-time Adaptive Reranking for Open-Set 3D Shape Retrieval

Large pretrained vision models have substantially improved appearance-based 3D shape retrieval, but they still confuse shapes that look similar while differing in geometry. Although geometry-aware features can reduce these errors, naive fusion of geometry and appearance may hurt retrieval when the two modalities are already well aligned. We propose GATE-3D, a lightweight query-adaptive reranking method that incorporates geometry without retraining the retrieval backbone. For each query, GATE-3D predicts how much a geometry-aware score should adjust the appearance-based ranking using features that capture disagreement between the two modalities. This selective design lets geometry contribute where it helps and stay silent where it would hurt. Experiments on three open-set 3D retrieval benchmarks show that GATE-3D improves over appearance-only retrieval and is more robust than always-on fusion. On the primary benchmark, it improves mAP@10 by 2.00 points over appearance-only retrieval (p=0.041); it also improves leave-one-category-out generalization and reduces geometric false positives by 10.8%. GATE-3D achieves competitive zero-shot results against DAC-based baselines. We further find that simple linear routing is more effective than a small MLP in the low-data regime, suggesting that cross-modal disagreement features matter more than model capacity for adaptive routing.

cs.CV↗
arXiv2026

CR-Refiner: An Object-Centric Optimal Transport Reranker for Edit-Conditioned 3D Scene Retrieval

Edit-conditioned 3D scene retrieval pairs a reference 3D room with a natural-language modification and retrieves rooms from a corpus that satisfy the edit. Three lines of prior work each fall short on this task. 2D composed image retrieval reasons over pixel-level edits and has no primitive for 3D object sets. 3D foundation encoders embed individual objects but cannot compose at the scene level. 3D scene-grounding methods localize references inside a static scene rather than rank modified rooms across a corpus. We present CR-Refiner, a training-free reranker that wraps any base retriever's top-K candidates with three components. A frozen LLM parses the edit into a structured query entity, and each candidate is scored by an unbalanced optimal-transport problem over a 1xG cost matrix coupling category, style, material, and geometry. The unbalanced solver lets the single-entity query drop mass on irrelevant objects, modelling the asymmetry directly. An axis-conditional structural prior adds size-keyword cues for geometric edits and subject-anchor direction cues for spatial edits. An LLM verifier refines the top three candidates with continuous confidence. Because no benchmark evaluates compositional matching over 3D object sets, we additionally release 3D-CER, 4,963 edit-conditioned queries over a 23,381-room indoor corpus across five edit axes, with multi-positive ground truth, CIRR-style hard subsets, and zero-target adversarials. Across three qualitatively distinct base retrievers, CR-Refiner consistently improves hard-subset R@1 and mAP@10 on every edit axis.

cs.CV↗
arXiv2026

RECEIPT: Deterministic, Reward-Hacking-Resistant Verification for White-Box Agentic XSS Discovery

Cross-Site Scripting (XSS) remains one of the most prevalent and damaging classes of web vulnerabilities. LLM-based coding agents offer a promising approach to XSS discovery by combining source-code reasoning with interactive testing against a running application. However, a coding agent's claims cannot be trusted on their own. We characterize three reward-hacking behaviors in white-box agentic XSS discovery and propose three requirements that an ideal verifier should meet. We present RECEIPT, a verification framework that makes agent-reported XSS findings trustworthy by enforcing environment isolation, PoC constraints, role separation, and verdict binding. Each confirmation therefore establishes two properties: the script runs in a real browser, and the payload was planted under the attacker role and executed in the victim role's browser. This constrained replay procedure makes validation deterministic and reproducible. We evaluate RECEIPT on 95 real-world web-application targets drawn from popular open-source projects. Within a $20 per-application budget, RECEIPT found 24 previously unknown XSS vulnerabilities, 12 of which have already been acknowledged by maintainers after responsible disclosure, and recovered the labeled CVE in 36% of known-vulnerability recovery targets. Compared with the same agent using self-judgment and with black-box scanners, RECEIPT confirms more real exploits while admitting no false positives.

cs.CR↗
arXiv2026

Emotion-based Recommender System

Recommender system is one of the most critical technologies for large internet companies such as Amazon and TikTok. Although millions of users use recommender systems globally everyday, and indeed, much data analysis work has been done to improve the technical accuracy of the system, to our limited knowledge, there has been little attention paid to analysis of users' emotion in recommender systems. In this paper, we create a new theory and metrics that could capture users' emotion when they are interacting with recommender systems. We also provide effective and efficient visualization techniques for visualization of users' emotion and its change in the customers' lifetime cycle. In the end, we design a framework for emotion-based recommendation algorithms, illustrated in a straightforward example with experimental results to demonstrate the effectiveness of our new theory.

cs.IR↗
arXiv2026

Beyond Unfolding: 60x Faster One-Stage Unmixing for Closely-Spaced Infrared Small Targets

Due to the optical diffraction limit and long imaging distances, Closely-Spaced Infrared Small Targets (CSIST) typically exhibit energy overlap, manifesting as indistinguishable blobs in infrared images. This ambiguity invalidates the one-to-one mapping assumption of traditional detection, thereby necessitating a paradigm shift towards CSIST Unmixing, which decomposes these blobs into discrete sub-targets. However, the dominant paradigm deep unfolding networks are shackled by the high latency and structural inflexibility intrinsic to their repetitively iterative architecture. To this end, we propose the Fast One-stage CSIST Unmixing Scheme (FOCUS), a one-stage lightweight paradigm which demonstrates that deep unfolding is not necessary. Motivated by the key observation that image super-resolution (SR) and CSIST Unmixing share an isomorphic degradation model, our insight is that it is possible to achieve a paradigm shift from image SR to CSIST Unmixing via completely transforming the label space, loss functions, and evaluation criteria. Specifically, to avoid entangling geometric recovery with artifact suppression, FOCUS adopts a single pass mapping with an internal coarse-to-fine flow that progressively refines target localization from coarse spatial distributions to finer sub-pixel precision. While sparsity regularization suppresses background clutter, it also attenuates target intensities. To compensate for this attenuation of valid signals, flux conservation is introduced as a competing constraint that restores signal energy back to target centers. To the best of our knowledge, this work is the first attempt to address this task via a lightweight one-stage framework without the DUN paradigm. Experiments demonstrate that our method matches or surpasses the state-of-the-art unfolding approaches in both localization and unmixing accuracy, while boosting the inference speed by 60x.

cs.CV↗
arXiv2026

Dark Energy in the DESI Era: A Brief Review of Evidence, Beyond-$Λ$CDM Interpretations, and Tensions

Recent baryon acoustic oscillation measurements from DESI provide important new clues for reassessing whether the standard $Λ$CDM model offers a sufficient description of the late-time expansion history of the Universe. When combined with cosmic microwave background and type Ia supernova data, these measurements show an apparent departure from the $Λ$CDM model, commonly described as dynamical dark energy (DDE) with equation of state crossing the phantom divide (i.e., quintom behavior). This review examines the current status of the DESI-motivated indications for DDE and their possible implications for physics beyond $Λ$CDM. We discuss how the strength of the preference for DDE depends on the adopted parametrization and dataset combination, and how residual systematics or internal tensions among datasets may affect its interpretation. At the background level, several mechanisms beyond $Λ$CDM can produce similar expansion histories. We therefore further discuss how the same effective departure from $w=-1$ may arise from physically distinct scenarios, including interacting dark energy, non-minimally coupled gravity, and non-standard dark matter. Meanwhile, these different new-physics interpretations may have different implications for current cosmological tensions, especially those involving $H_0$, $S_8$, and $\sum m_ν$. In conclusion, the question posed by DESI is not merely whether dark energy evolves with time, but rather how, within the framework of precision cosmology, to disentangle new physics scenarios from systematic errors.

astro-ph.CO↗
arXiv2026

Entropy-Preserving Supervised Fine-Tuning via Adaptive Self-Distillation for Large Reasoning Models

The standard post-training recipe for large reasoning models, supervised fine-tuning followed by reinforcement learning (SFT-then-RL), may limit the benefits of the RL stage: while SFT imitates expert demonstrations, it often causes overconfidence and reduces generation diversity, leaving RL with a narrowed solution space to explore. Adding entropy regularization during SFT is not a cure-all; it tends to flatten token distributions toward uniformity, increasing entropy without improving meaningful exploration capability. In this paper, we propose CurioSFT, an entropy-preserving SFT method designed to enhance exploration capabilities through intrinsic curiosity. It consists of (a) Self-Exploratory Distillation, which distills the model toward a self-generated, temperature-scaled teacher to encourage exploration within its capability; and (b) Entropy-Guided Temperature Selection, which adaptively adjusts distillation strength to mitigate knowledge forgetting by amplifying exploration at reasoning tokens while stabilizing factual tokens. Extensive experiments on mathematical reasoning tasks demonstrate that, in SFT stage, CurioSFT outperforms the vanilla SFT by 2.5 points on in-distribution tasks and 2.9 points on out-of-distribution tasks. We also verify that exploration capabilities preserved during SFT successfully translate into concrete gains in RL stage, yielding an average improvement of 5.0 points.

cs.LG↗
arXiv2026

Layer-Parallel Inference Reduces Encrypted Nonlinear Depth in Transformers

Fully homomorphic encryption (FHE) enables computation on encrypted data, but practical encrypted Transformer inference is bottlenecked by the sequential composition of many nonlinear blocks. We study whether Structured Newton Layer Parallelism (SNLP) can make this inter-layer composition more FHE-friendly: each Transformer block still requires polynomial approximations for operations such as softmax and RMSNorm, but SNLP reduces the layerwise sequential nonlinear depth from L stages to a small number of solver iterations plus linear structured corrections. Using a simulation framework based on Chebyshev polynomial approximations, we measure error accumulation under sequential versus SNLP inference across 8 models and 4 architecture families. On a 0.5B IDN-trained model, SNLP reduces symbolic bootstraps from 53 to 20 (2.65x) with only +1.2% perplexity degradation, while lowering error amplification (1.36x vs. 1.42x). Across all tested models, SNLP has lower amplification than sequential inference. Ablations show that softmax approximation dominates the error budget and CKKS arithmetic noise is negligible in our setting, suggesting that SNLP is complementary to block-level FHE-friendly operator design rather than a replacement for it.

cs.LG↗
arXiv2026

LSTrans: Efficient Knowledge Transfer for Lightweight and Automated ECG Classification

Deploying deep learning models for automated electrocardiogram classification on resource-constrained wearable devices remains challenging due to high computational costs. To address this, we propose LSTrans, a lightweight hybrid model designed for efficient and sensitive ECG analysis. LSTrans introduces a specialized 1D convolutional backbone with an interleaved layer architecture to capture both macroscopic rhythmic trends and microscopic morphological variations. This backbone is cascaded with a Transformer encoder to model long-range temporal dependencies, incorporating Low-Rank Adaptation across critical layers to compress the model and reduce the trainable parameter space. We further employ homogeneous and heterogeneous knowledge distillation to transfer diagnostic expertise from high-capacity teacher models to the student. Experimental results on multiple benchmark datasets demonstrate that LSTrans achieves a competitive balance between diagnostic sensitivity and resource efficiency, substantially reducing peak memory footprints and training latency during downstream adaptation. The source code is available for review at https://github.com/zyee00128/LSTrans4BIBM.

cs.LG↗