Searcharxiv⌕ Search

arXiv subjects

Lei Wang

Publications and source records attributed to Lei Wang.

At least 199 records · Page 11Linked to original sources
arXiv2026

QaRL: Rollout-Aligned Quantization-Aware RL for Fast and Stable Training under Training--Inference Mismatch

Large language model (LLM) reinforcement learning (RL) pipelines are often bottlenecked by rollout generation, making end-to-end training slow. Recent work mitigates this by running rollouts with quantization to accelerate decoding, which is the most expensive stage of the RL loop. However, these setups destabilize optimization by amplifying the training-inference gap: rollouts are operated at low precision, while learning updates are computed at full precision. To address this challenge, we propose QaRL (Rollout Alignment Quantization-Aware RL), which aligns training-side forward with the quantized rollout to minimize mismatch. We further identify a failure mode in quantized rollouts: long-form responses tend to produce repetitive, garbled tokens (error tokens). To mitigate these problems, we introduce TBPO (Trust-Band Policy Optimization), a sequence-level objective with dual clipping for negative samples, aimed at keeping updates within the trust region. On Qwen3-30B-A3B MoE for math problems, QaRL outperforms quantized-rollout training by +5.5 while improving stability and preserving low-bit throughput benefits.

cs.LG↗
arXiv2026

Precise $^{136}$Xe Double Beta Decay Measurement in PandaX-4T with Implications on the Nuclear Matrix Elements and Majorons

The continuous spectrum of double beta decay ($ββ$) provides a sensitive probe to test the predictions of the standard model and to search for signatures of new physics beyond it. We present a comprehensive analysis of the $^{136}$Xe $ββ$ spectrum utilizing $39.1 \pm 0.7~\textrm{kg}\cdot\textrm{yr}$ of $^{136}$Xe exposure from the PandaX-4T experiment. The analysis yields the most precise measurement to date of the $^{136}$Xe two-neutrino double beta decay ($2νββ$) half-life, $(2.14 \pm 0.05) \times 10^{21}$ years, the uncertainty of which is reduced by a factor of 2 compared to our previous result. We measure the parameter $ξ_{31}^{2ν}$, defined as the ratio between the subleading and leading components of the $^{136}$Xe $2νββ$ nuclear matrix element, to be $0.59^{+0.41}_{-0.38}$, which is consistent with theoretical predictions. We also search for Majoron-emitting modes of $^{136}$Xe $ββ$, establishing the most stringent limit for the spectral index $n=7$.

nucl-ex↗
arXiv2026

MARS: Enabling Autoregressive Models Multi-Token Generation

Autoregressive (AR) language models generate text one token at a time, even when consecutive tokens are highly predictable given earlier context. We introduce MARS (Mask AutoRegreSsion), a lightweight fine-tuning method that teaches an instruction-tuned AR model to predict multiple tokens per forward pass. MARS adds no architectural modifications, no extra parameters, and produces a single model that can still be called exactly like the original AR model with no performance degradation. Unlike speculative decoding, which maintains a separate draft model alongside the target, or multi-head approaches such as Medusa, which attach additional prediction heads, MARS requires only continued training on existing instruction data. When generating one token per forward pass, MARS matches or exceeds the AR baseline on six standard benchmarks. When allowed to accept multiple tokens per step, it maintains baseline-level accuracy while achieving 1.5-1.7x throughput. We further develop a block-level KV caching strategy for batch inference, achieving up to 1.71x wall-clock speedup over AR with KV cache on Qwen2.5-7B. Finally, MARS supports real-time speed adjustment via confidence thresholding: under high request load, the serving system can increase throughput on the fly without swapping models or restarting, providing a practical latency-quality knob for deployment.

cs.CL↗
arXiv2026

Biologically Inspired Event-Based Perception and Sample-Efficient Learning for High-Speed Table Tennis Robots

Perception and decision-making in high-speed dynamic scenarios remain challenging for current robots. In contrast, humans and animals can rapidly perceive and make decisions in such environments. Taking table tennis as a typical example, conventional frame-based vision sensors suffer from motion blur, high latency and data redundancy, which can hardly meet real-time, accurate perception requirements. Inspired by the human visual system, event-based perception methods address these limitations through asynchronous sensing, high temporal resolution, and inherently sparse data representations. However, current event-based methods are still restricted to simplified, unrealistic ball-only scenarios. Meanwhile, existing decision-making approaches typically require thousands of interactions with the environment to converge, resulting in significant computational costs. In this work, we present a biologically inspired approach for high-speed table tennis robots, combining event-based perception with sample-efficient learning. On the perception side, we propose an event-based ball detection method that leverages motion cues and geometric consistency, operating directly on asynchronous event streams without frame reconstruction, to achieve robust and efficient detection in real-world rallies. On the decision-making side, we introduce a human-inspired, sample-efficient training strategy that first trains policies in low-speed scenarios, progressively acquiring skills from basic to advanced, and then adapts them to high-speed scenarios, guided by a case-dependent temporally adaptive reward and a reward-threshold mechanism. With the same training episodes, our method improves return-to-target accuracy by 35.8%. These results demonstrate the effectiveness of biologically inspired perception and decision-making for high-speed robotic systems.

cs.RO↗
arXiv2026

LLM Agents as Social Scientists: A Human-AI Collaborative Platform for Social Science Automation

Traditional social science research often requires designing complex experiments across vast methodological spaces and depends on real human participants, making it labor-intensive, costly, and difficult to scale. Here we present S-Researcher, an LLM-agent-based platform that assists researchers in conducting social science research more efficiently and at greater scale by "siliconizing" both the research process and the participant pool. To build S-Researcher, we first develop YuLan-OneSim, a large-scale social simulation system designed around three core requirements: generality via auto-programming from natural language to executable scenarios, scalability via a distributed architecture supporting up to 100,000 concurrent agents, and reliability via feedback-driven LLM fine-tuning. Leveraging this system, S-Researcher supports researchers in designing social experiments, simulating human behavior with LLM agents, analyzing results, and generating reports, forming a complete human-AI collaborative research loop in which researchers retain oversight and intervention at every stage. We operationalize LLM simulation research paradigms into three canonical reasoning modes (induction, deduction, and abduction) and validate S-Researcher through systematic case studies: inductive reproduction of cultural dynamics consistent with Axelrod's theory, deductive testing of competing hypotheses on teacher attention validated against survey data, and abductive identification of a cooperation mechanism in public goods games confirmed by human experiments. S-Researcher establishes a new human--AI collaborative paradigm for social science, in which computational simulation augments human researchers to accelerate discovery across the full spectrum of social inquiry.

cs.AI↗
arXiv2026

A Cross-graph Tuning-free GNN Prompting Framework

GNN prompting aims to adapt models across tasks and graphs without requiring extensive retraining. However, most existing graph prompt methods still require task-specific parameter updates and face the issue of generalizing across graphs, limiting their performance and undermining the core promise of prompting. In this work, we introduce a Cross-graph Tuning-free Prompting Framework (CTP), which supports both homogeneous and heterogeneous graphs, can be directly deployed to unseen graphs without further parameter tuning, and thus enables a plug-and-play GNN inference engine. Extensive experiments on few-shot prediction tasks show that, compared to SOTAs, CTP achieves an average accuracy gain of 30.8% and a maximum gain of 54%, confirming its effectiveness and offering a new perspective on graph prompt learning.

cs.LG↗
arXiv2026

Uncover the correlation between jet energy correlators and multiplicity fluctuations

The energy-energy correlator (EEC) and multiplicity are two fundamental observables probing complementary aspects of QCD jets: the former characterizes the angular structure of energy flows in a scale-dependent manner, while the latter is sensitive to the entire history of particle production. In this \emph{Letter}, we uncover a nontrivial correlation between them by studying the EEC as a function of jet internal multiplicity. We introduce the multiplicity-conditioned EEC jet function (MCJF) and perform a factorization calculation to next-to-leading order accuracy. It is found that, for jet samples selected at a given normalized multiplicity $ν= N_{\rm ch}/\langle N_{\rm ch} \rangle$, the EEC in the angular region $Λ_{\rm QCD}/p_{T,\rm jet}\llχ\ll R$ acquires a $ν$-dependent anomalous dimension. Thus the $ν$-conditioned EEC provides a direct and robust probe to the multiplicity generating function in the perturbative regime. In addition, understanding $ν$ dependence of the EEC is also crucial for isolating possible multiplicity-dependent bias effects in the EEC measurements in nuclear environment.

hep-ph↗
arXiv2026

The Overlooked Repetitive Lengthening Form in Sentiment Analysis

Individuals engaging in online communication frequently express personal opinions with informal styles (e.g., memes and emojis). While Language Models (LMs) with informal communications have been widely discussed, a unique and emphatic style, the Repetitive Lengthening Form (RLF), has been overlooked for years. In this paper, we explore answers to two research questions: 1) Is RLF important for sentiment analysis (SA)? 2) Can LMs understand RLF? Inspired by previous linguistic research, we curate \textbf{Lengthening}, the first multi-domain dataset with 850k samples focused on RLF for SA. Moreover, we introduce \textbf{Exp}lainable \textbf{Instruct}ion Tuning (\textbf{ExpInstruct}), a two-stage instruction tuning framework aimed to improve both performance and explainability of LLMs for RLF. We further propose a novel unified approach to quantify LMs' understanding of informal expressions. We show that RLF sentences are expressive expressions and can serve as signatures of document-level sentiment. Additionally, RLF has potential value for online content analysis. Our results show that fine-tuned Pre-trained Language Models (PLMs) can surpass zero-shot GPT-4 in performance but not in explanation for RLF. Finally, we show ExpInstruct can improve the open-sourced LLMs to match zero-shot GPT-4 in performance and explainability for RLF with limited samples. Code and sample data are available at https://github.com/Tom-Owl/OverlookedRLF

cs.CL↗
arXiv2026

MiroEval: Benchmarking Multimodal Deep Research Agents in Process and Outcome

Recent progress in deep research systems has been impressive, but evaluation still lags behind real user needs. Existing benchmarks predominantly assess final reports using fixed rubrics, failing to evaluate the underlying research process. Most also offer limited multimodal coverage, rely on synthetic tasks that do not reflect real-world query complexity, and cannot be refreshed as knowledge evolves. To address these gaps, we introduce MiroEval, a benchmark and evaluation framework for deep research systems. The benchmark comprises 100 tasks (70 text-only, 30 multimodal), all grounded in real user needs and constructed via a dual-path pipeline that supports periodic updates, enabling a live and evolving setting. The proposed evaluation suite assesses deep research systems along three complementary dimensions: adaptive synthesis quality evaluation with task-specific rubrics, agentic factuality verification via active retrieval and reasoning over both web sources and multimodal attachments, and process-centric evaluation audits how the system searches, reasons, and refines throughout its investigation. Evaluation across 13 systems yields three principal findings: the three evaluation dimensions capture complementary aspects of system capability, with each revealing distinct strengths and weaknesses across systems; process quality serves as a reliable predictor of overall outcome while revealing weaknesses invisible to output-level metrics; and multimodal tasks pose substantially greater challenges, with most systems declining by 3 to 10 points. The MiroThinker series achieves the most balanced performance, with MiroThinker-H1 ranking the highest overall in both settings. Human verification and robustness results confirm the reliability of the benchmark and evaluation framework. MiroEval provides a holistic diagnostic tool for the next generation of deep research agents.

cs.AI↗
arXiv2026

Probing a long-lived pseudoscalar in type-I 2HDM with displaced vertices and jets at the LHC

In the type-I two-Higgs-doublet model, the pseudoscalar $A$ can act as a long-lived particle (LLP) for sufficiently large values of $\tanβ$. At the LHC, the $A$ particles are predominantly produced in pairs through $pp \to W^*/Z^* \to H^\pm/H \, A$, with subsequent decays $H^{\pm}/H \to W^\pm/Z\, A$. For the mass range of our interest, $10~\text{GeV}\lesssim m_A \lesssim 100~\text{GeV}$, the pseudoscalar $A$ typically decays into a pair of bottom quarks after traveling a macroscopic distance from its production point, giving rise to displaced-vertex (DV) signatures inside the inner detector. We perform Monte Carlo simulations of signal events with DVs plus jets, and assess the discovery prospects of $A$ as an LLP at the ATLAS and CMS experiments. Our findings show that a substantial portion of the parameter space with $m_A>10$ GeV has already been excluded by LHC Run-2 data, while the high-luminosity LHC will be able to probe broader regions.

hep-ph↗
arXiv2026

A Nonlinear Incremental Approach for Replay Attack Detection

Replay attacks comprise replaying previously recorded sensor measurements and injecting malicious signals into a physical plant, causing great damage to cyber-physical systems. Replay attack detection has been widely studied for linear systems, whereas limited research has been reported for nonlinear cases. In this paper, the replay attack is studied in the context of a nonlinear plant controlled by an observer-based output feedback controller. We first analyze replay attack detection using an innovation-based detector and reveal that this detector alone may fail to detect such attacks. Consequently, we turn to a watermark-based design framework to improve the detection. In the proposed framework, the effects of the watermark on attack detection and closed-loop system performance loss are quantified by two indices, which exploit the incremental gains of nonlinear systems. To balance the detection performance and control system performance loss, an explicit optimization problem is formulated. Moreover, to achieve a better balance, we generalize the proposed watermark design framework to co-design the watermark, controller and observer. Numerical simulations are presented to validate the proposed frameworks.

eess.SY↗
arXiv2026

A Global Spacetime Optimization Approach to the Real-Space Time-Dependent Schrödinger Equation

The time-dependent Schrödinger equation (TDSE) in real space is fundamental to understanding the dynamics of many-electron quantum systems, with applications ranging from quantum chemistry to condensed matter physics and materials science. However, solving the TDSE for complex fermionic systems remains a significant challenge, particularly due to the need to capture the time-evolving many-body correlations, while the antisymmetric nature of fermionic wavefunctions complicates the function space in which these solutions must be represented. We propose a general-purpose neural network framework for solving the real-space TDSE, Fermionic Antisymmetric Spatio-Temporal Network, which treats time as an explicit input alongside spatial coordinates, enabling a unified spatiotemporal representation of complex, antisymmetric wavefunctions for fermionic systems. This approach formulates the TDSE as a global optimization problem, avoiding step-by-step propagation and supporting highly parallelizable training. The method is demonstrated on five benchmark problems, achieving excellent agreement with reference solutions across all cases. These results demonstrate the method's accuracy and flexibility within the bound-state manifold across various dimensions and interaction regimes. While the current localized Ansatz inherently restricts the description of extensive ionization and continuum states, the method demonstrates the capability to stably simulate coherent multi-electron dynamics over extended time windows. Our framework offers a highly expressive alternative to traditional basis-dependent or mean-field methods, opening new possibilities for ab initio simulations of time-dependent quantum systems, with applications in quantum dynamics, molecular control, and ultrafast spectroscopy.

quant-ph↗
arXiv2026

EmoTaG: Emotion-Aware Talking Head Synthesis on Gaussian Splatting with Few-Shot Personalization

Audio-driven 3D talking head synthesis has advanced rapidly with Neural Radiance Fields (NeRF) and 3D Gaussian Splatting (3DGS). By leveraging rich pre-trained priors, few-shot methods enable instant personalization from just a few seconds of video. However, under expressive facial motion, existing few-shot approaches often suffer from geometric instability and audio-emotion mismatch, highlighting the need for more effective emotion-aware motion modeling. In this work, we present EmoTaG, a few-shot emotion-aware 3D talking head synthesis framework built on the Pretrain-and-Adapt paradigm. Our key insight is to reformulate motion prediction in a structured FLAME parameter space rather than directly deforming 3D Gaussians, thereby introducing explicit geometric priors that improve motion stability. Building upon this, we propose a Gated Residual Motion Network (GRMN), which captures emotional prosody from audio while supplementing head pose and upper-face cues absent from audio, enabling expressive and coherent motion generation. Extensive experiments demonstrate that EmoTaG achieves state-of-the-art performance in emotional expressiveness, lip synchronization, visual realism, and motion stability.

cs.CV↗
arXiv2026

An End-to-end Flight Control Network for High-speed UAV Obstacle Avoidance based on Event-Depth Fusion

Achieving safe, high-speed autonomous flight in complex environments with static, dynamic, or mixed obstacles remains challenging, as a single perception modality is incomplete. Depth cameras are effective for static objects but suffer from motion blur at high speeds. Conversely, event cameras excel at capturing rapid motion but struggle to perceive static scenes. To exploit the complementary strengths of both sensors, we propose an end-to-end flight control network that achieves feature-level fusion of depth images and event data through a bidirectional crossattention module. The end-to-end network is trained via imitation learning, which relies on high-quality supervision. Building on this insight, we design an efficient expert planner using Spherical Principal Search (SPS). This planner reduces computational complexity from $O(n^2)$ to $O(n)$ while generating smoother trajectories, achieving over 80% success rate at 17m/s--nearly 20% higher than traditional planners. Simulation experiments show that our method attains a 70-80% success rate at 17 m/s across varied scenes, surpassing single-modality and unidirectional fusion models by 10-20%. These results demonstrate that bidirectional fusion effectively integrates event and depth information, enabling more reliable obstacle avoidance in complex environments with both static and dynamic objects.

cs.RO↗
arXiv2026

IncreRTL: Traceability-Guided Incremental RTL Generation under Requirement Evolution

Large language models (LLMs) have shown promise in generating RTL code from natural-language descriptions, but existing methods remain static and struggle to adapt to evolving design requirements, potentially causing structural drift and costly full regeneration. We propose IncreRTL, a LLM-driven framework for incremental RTL generation under requirement evolution. By constructing requirement-code traceability links to locate and regenerate affected code segments, IncreRTL achieves accurate and consistent updates. Evaluated on our newly constructed EvoRTL-Bench, IncreRTL demonstrates notable improvements in regeneration consistency and efficiency, advancing LLM-based RTL generation toward practical engineering deployment.

cs.SE↗
arXiv2026

Explicit Model Predictive Control with Quantum Encryption

This paper studies quantum-encrypted explicit MPC for constrained discrete-time linear systems in a cloud-based architecture. A finite-horizon quadratic MPC problem is solved offline to obtain a piecewise-affine controller. Shared quantum keys generated from Bell pairs and protected by quantum key distribution are used to encrypt the online control evaluation between the sensor and actuator. Based on this architecture, we develop a lightweight encrypted explicit MPC protocol, prove exact recovery of the plaintext control action, and characterize its computational efficiency. Numerical results demonstrate lower online complexity than classical encrypted MPC, while security is discussed in terms of confidentiality of plant data and control inputs.

eess.SY↗
arXiv2026

Efficient photo-Nernst terahertz emission in single heavy-metal films

State-of-the-art metallic terahertz (THz) emitters rely predominantly on spintronic heterostructures, where heavy metals serve as passive spin-to-charge converters. Here, we demonstrate efficient THz radiation from standalone Pt nanofilms at cryogenic temperatures and under external magnetic fields. The governing mechanism is identified as the ultrafast photo-Nernst effect, wherein a transient thermal gradient drives a transverse charge current. The THz emission polarity is directly dictated by the sign of the Nernst coefficient, as verified by the phase reversal observed between Pt and W or Ta. Remarkably, both thickness scaling and alloying-induced suppression of thermal conductivity independently amplify the single-layer emission to levels comparable with benchmark spintronic bilayers. These findings redefine the established role of heavy metals from passive spin-sinks to active THz emitters, uncovering a universal emission paradigm applicable across diverse spintronic and quantum materials.

cond-mat.mes-hall↗
arXiv2026

Subspace Kernel Learning on Tensor Sequences

Learning from structured multi-way data, represented as higher-order tensors, requires capturing complex interactions across tensor modes while remaining computationally efficient. We introduce Uncertainty-driven Kernel Tensor Learning (UKTL), a novel kernel framework for $M$-mode tensors that compares mode-wise subspaces derived from tensor unfoldings, enabling expressive and robust similarity measure. To handle large-scale tensor data, we propose a scalable Nyström kernel linearization with dynamically learned pivot tensors obtained via soft $k$-means clustering. A key innovation of UKTL is its uncertainty-aware subspace weighting, which adaptively down-weights unreliable mode components based on estimated confidence, improving robustness and interpretability in comparisons between input and pivot tensors. Our framework is fully end-to-end trainable and naturally incorporates both multi-way and multi-mode interactions through structured kernel compositions. Extensive evaluations on action recognition benchmarks (NTU-60, NTU-120, Kinetics-Skeleton) show that UKTL achieves state-of-the-art performance, superior generalization, and meaningful mode-wise insights. This work establishes a principled, scalable, and interpretable kernel learning paradigm for structured multi-way and multi-modal tensor sequences.

cs.LG↗