arXiv subjects

Shuai Nie

Publications and source records attributed to Shuai Nie.

At least 19 recordsLinked to original sources

arXiv2026

From Profiles to Steering Vectors: Global Sparse Priors and Local Semantic Calibration for Personalized Text Generation

Personalized text generation requires models to capture user-specific writing styles from historical data. Existing approaches based on retrieval, parameter-efficient fine-tuning, or activation steering either introduce inference and storage overhead or struggle to separate stylistic signals from semantic content. We propose GLASS, a training-free framework for personalized generation via Global-Local Activation Steering with Sparse priors. GLASS uses sparse autoencoders to extract a global user-style prior from historical responses and constructs local contrastive style vectors over clustered interaction scenarios. During inference, it jointly injects global and local vectors into different model layers, enabling context-aware personalization without retrieval or parameter updates. Experiments on LaMP and LongLaMP show that GLASS outperforms retrieval-, fine-tuning-, and steering-based baselines across ROUGE metrics and LLM-as-judge evaluations. Further analyses show that SAE-based representations are more robust to topic and length shifts, suggesting better disentanglement of stylistic information from semantic residue.

cs.AI↗

arXiv2026

ProactiveMobile: A Comprehensive Benchmark for Boosting Proactive Intelligence on Mobile Devices

Multimodal large language models (MLLMs) have made significant progress in mobile agent development, yet their capabilities are predominantly confined to a reactive paradigm, where they merely execute explicit user commands. The emerging paradigm of proactive intelligence, where agents autonomously anticipate needs and initiate actions, represents the next frontier for mobile agents. However, its development is critically bottlenecked by the lack of benchmarks that can address real-world complexity and enable objective, executable evaluation. To overcome these challenges, we introduce ProactiveMobile, a comprehensive benchmark designed to systematically advance research in this domain. ProactiveMobile formalizes the proactive task as inferring latent user intent across four dimensions of on-device contextual signals and generating an executable function sequence from a comprehensive function pool of 63 APIs. The benchmark features over 3,660 instances of 14 scenarios that embrace real-world complexity through multi-answer annotations. To ensure quality, a team of 30 experts conducts a final audit of the benchmark, verifying factual accuracy, logical consistency, and action feasibility, and correcting any non-compliant entries. Extensive experiments demonstrate that our fine-tuned Qwen2.5-VL-7B-Instruct achieves a success rate of 19.15%, outperforming o1 (15.71%) and GPT-5 (7.39%). This result indicates that proactivity is a critical competency widely lacking in current MLLMs, yet it is learnable, emphasizing the importance of the proposed benchmark for proactivity evaluation.

cs.AI↗

arXiv2026

ProUIE: A Macro-to-Micro Progressive Learning Method for LLM-based Universal Information Extraction

LLM-based universal information extraction (UIE) methods often rely on additional information beyond the original training data, which increases training complexity yet often yields limited gains. To address this, we propose ProUIE, a Macro-to-Micro progressive learning approach that improves UIE without introducing any external information. ProUIE consists of three stages: (i) macro-level Complete Modeling (CM), which learns NER, RE, and EE along their intrinsic difficulty order on the full training data to build a unified extraction foundation, (ii) meso-level Streamlined Alignment (SA), which operates on sampled data with simplified target formats, streamlining and regularizing structured outputs to make them more concise and controllable, and (iii) micro-level Deep Exploration (DE), which applies GRPO with stepwise fine-grained rewards (SFR) over structural units to guide exploration and improve performance. Experiments on 36 public datasets show that ProUIE consistently improves unified extraction, outperforming strong instruction-tuned baselines on average for NER and RE while using a smaller backbone, and it further demonstrates clear gains in large-scale production-oriented information extraction.

cs.CL↗

arXiv2026

SSL: Sweet Spot Learning for Differentiated Guidance in Agentic Optimization

Reinforcement learning with verifiable rewards has emerged as a powerful paradigm for training intelligent agents. However, existing methods typically employ binary rewards that fail to capture quality differences among trajectories achieving identical outcomes, thereby overlooking potential diversity within the solution space. Inspired by the ``sweet spot'' concept in tennis-the racket's core region that produces optimal hitting effects, we introduce \textbf{S}weet \textbf{S}pot \textbf{L}earning (\textbf{SSL}), a novel framework that provides differentiated guidance for agent optimization. SSL follows a simple yet effective principle: progressively amplified, tiered rewards guide policies toward the sweet-spot region of the solution space. This principle naturally adapts across diverse tasks: visual perception tasks leverage distance-tiered modeling to reward proximity, while complex reasoning tasks reward incremental progress toward promising solutions. We theoretically demonstrate that SSL preserves optimal solution ordering and enhances the gradient signal-to-noise ratio, thereby fostering more directed optimization. Extensive experiments across GUI perception, short/long-term planning, and complex reasoning tasks show consistent improvements over strong baselines on 12 benchmarks, achieving up to 2.5X sample efficiency gains and effective cross-task transferability. Our work establishes SSL as a general principle for training capable and robust agents.

cs.CL↗

arXiv2026

Unified Multimodal and Multilingual Retrieval via Multi-Task Learning with NLU Integration

Multimodal retrieval systems typically employ Vision Language Models (VLMs) that encode images and text independently into vectors within a shared embedding space. Despite incorporating text encoders, VLMs consistently underperform specialized text models on text-only retrieval tasks. Moreover, introducing additional text encoders increases storage, inference overhead, and exacerbates retrieval inefficiencies, especially in multilingual settings. To address these limitations, we propose a multi-task learning framework that unifies the feature representation across images, long and short texts, and intent-rich queries. To our knowledge, this is the first work to jointly optimize multilingual image retrieval, text retrieval, and natural language understanding (NLU) tasks within a single framework. Our approach integrates image and text retrieval with a shared text encoder that is enhanced by NLU features for intent understanding and retrieval accuracy.

cs.IR↗

arXiv2025

HyperVL: An Efficient and Dynamic Multimodal Large Language Model for Edge Devices

Current multimodal large lanauge models possess strong perceptual and reasoning capabilities, however high computational and memory requirements make them difficult to deploy directly on on-device environments. While small-parameter models are progressively endowed with strong general capabilities, standard Vision Transformer (ViT) encoders remain a critical bottleneck, suffering from excessive latency and memory consumption when processing high-resolution inputs.To address these challenges, we introduce HyperVL, an efficient multimodal large language model tailored for on-device inference. HyperVL adopts an image-tiling strategy to cap peak memory usage and incorporates two novel techniques: (1) a Visual Resolution Compressor (VRC) that adaptively predicts optimal encoding resolutions to eliminate redundant computation, and (2) Dual Consistency Learning (DCL), which aligns multi-scale ViT encoders within a unified framework, enabling dynamic switching between visual branches under a shared LLM. Extensive experiments demonstrate that HyperVL achieves state-of-the-art performance among models of comparable size across multiple benchmarks. Furthermore, it significantly significantly reduces latency and power consumption on real mobile devices, demonstrating its practicality for on-device multimodal inference.

cs.CV↗

arXiv2025

From Imitation to Discrimination: Toward A Generalized Curriculum Advantage Mechanism Enhancing Cross-Domain Reasoning Tasks

Reinforcement learning has emerged as a paradigm for post-training large language models, boosting their reasoning capabilities. Such approaches compute an advantage value for each sample, reflecting better or worse performance than expected, thereby yielding both positive and negative signals for training. However, the indiscriminate mixing of the two signals in existing methods, especially from the early stages, may lead to ambiguous guidance and limited gains. To address this issue, we propose **CAPO** (**C**urriculum **A**dvantage **P**olicy **O**ptimization), an adaptive curriculum mechanism based on advantage signals. The proposed mechanism bootstraps imitation learning with positive-only advantage samples to establish robust foundations, and subsequently introduces negative signals to cultivate discriminative capabilities, thereby improving generalization across complex scenarios. Compatible with diverse optimization methods including GRPO, PPO, RLOO, and Reinforce++, our method consistently achieves stable and significant improvements in mathematical reasoning tasks, and further generalizes effectively to multimodal Graphical User Interface (GUI) reasoning scenarios, establishing itself as a versatile and robust optimization framework.

cs.CL↗

arXiv2025

Spectrum Opportunities for the Wireless Future: From Direct-to-Device Satellite Applications to 6G Cellular

For next-generation wireless networks, both the upper mid-band and terahertz spectra are gaining global attention. This article provides an in-depth analysis of recent regulatory rulings and spectrum preferences issued by international standard bodies. We highlight promising frequency bands earmarked for 6G and beyond, and offer examples that illuminate the passive service protections and spectrum feasibility for coexistence between terrestrial and non-terrestrial wireless networks.

eess.SY↗

arXiv2024

ADD 2022: the First Audio Deep Synthesis Detection Challenge

Audio deepfake detection is an emerging topic, which was included in the ASVspoof 2021. However, the recent shared tasks have not covered many real-life and challenging scenarios. The first Audio Deep synthesis Detection challenge (ADD) was motivated to fill in the gap. The ADD 2022 includes three tracks: low-quality fake audio detection (LF), partially fake audio detection (PF) and audio fake game (FG). The LF track focuses on dealing with bona fide and fully fake utterances with various real-world noises etc. The PF track aims to distinguish the partially fake audio from the real. The FG track is a rivalry game, which includes two tasks: an audio generation task and an audio fake detection task. In this paper, we describe the datasets, evaluation metrics, and protocols. We also report major findings that reflect the recent advances in audio deepfake detection tasks.

cs.SD↗

arXiv2024

Can large language models understand uncommon meanings of common words?

Large language models (LLMs) like ChatGPT have shown significant advancements across diverse natural language understanding (NLU) tasks, including intelligent dialogue and autonomous agents. Yet, lacking widely acknowledged testing mechanisms, answering `whether LLMs are stochastic parrots or genuinely comprehend the world' remains unclear, fostering numerous studies and sparking heated debates. Prevailing research mainly focuses on surface-level NLU, neglecting fine-grained explorations. However, such explorations are crucial for understanding their unique comprehension mechanisms, aligning with human cognition, and finally enhancing LLMs' general NLU capacities. To address this gap, our study delves into LLMs' nuanced semantic comprehension capabilities, particularly regarding common words with uncommon meanings. The idea stems from foundational principles of human communication within psychology, which underscore accurate shared understandings of word semantics. Specifically, this paper presents the innovative construction of a Lexical Semantic Comprehension (LeSC) dataset with novel evaluation metrics, the first benchmark encompassing both fine-grained and cross-lingual dimensions. Introducing models of both open-source and closed-source, varied scales and architectures, our extensive empirical experiments demonstrate the inferior performance of existing models in this basic lexical-meaning understanding task. Notably, even the state-of-the-art LLMs GPT-4 and GPT-3.5 lag behind 16-year-old humans by 3.9% and 22.3%, respectively. Additionally, multiple advanced prompting techniques and retrieval-augmented generation are also introduced to help alleviate this trouble, yet limitations persist. By highlighting the above critical shortcomings, this research motivates further investigation and offers novel insights for developing more intelligent LLMs.

cs.CL↗

arXiv2024

KS-LLM: Knowledge Selection of Large Language Models with Evidence Document for Question Answering

Large language models (LLMs) suffer from the hallucination problem and face significant challenges when applied to knowledge-intensive tasks. A promising approach is to leverage evidence documents as extra supporting knowledge, which can be obtained through retrieval or generation. However, existing methods directly leverage the entire contents of the evidence document, which may introduce noise information and impair the performance of large language models. To tackle this problem, we propose a novel Knowledge Selection of Large Language Models (KS-LLM) method, aiming to identify valuable information from evidence documents. The KS-LLM approach utilizes triples to effectively select knowledge snippets from evidence documents that are beneficial to answering questions. Specifically, we first generate triples based on the input question, then select the evidence sentences most similar to triples from the evidence document, and finally combine the evidence sentences and triples to assist large language models in generating answers. Experimental comparisons on several question answering datasets, such as TriviaQA, WebQ, and NQ, demonstrate that the proposed method surpasses the baselines and achieves the best results.

cs.CL↗

arXiv2023

ADD 2023: the Second Audio Deepfake Detection Challenge

Audio deepfake detection is an emerging topic in the artificial intelligence community. The second Audio Deepfake Detection Challenge (ADD 2023) aims to spur researchers around the world to build new innovative technologies that can further accelerate and foster research on detecting and analyzing deepfake speech utterances. Different from previous challenges (e.g. ADD 2022), ADD 2023 focuses on surpassing the constraints of binary real/fake classification, and actually localizing the manipulated intervals in a partially fake speech as well as pinpointing the source responsible for generating any fake audio. Furthermore, ADD 2023 includes more rounds of evaluation for the fake audio game sub-challenge. The ADD 2023 challenge includes three subchallenges: audio fake game (FG), manipulation region location (RL) and deepfake algorithm recognition (AR). This paper describes the datasets, evaluation metrics, and protocols. Some findings are also reported in audio deepfake detection tasks.

cs.SD↗

arXiv2022

TeraHertz Band Communication: An Old Problem Revisited and Research Directions for the Next Decade

Terahertz (THz) band communications are envisioned as a key technology for 6G and Beyond. As a fundamental wireless infrastructure, THz communication can boost abundant promising applications. In 2014, our team published two comprehensive roadmaps for the development and progress of THz communication networks [1], [2], which helped the research community to start research on this subject afterwards. The topic of THz communications became very important and appealing to the research community due to 6G wireless systems design and development in recent years. Many papers are getting published covering different aspects of wireless systems using the THz band. With this paper, our aim is looking back to the last decade and revisiting the old problems and pointing out what has been achieved in the research community so far. Furthermore, in this paper, open challenges and new research directions still to be investigated for the THz band communication systems are presented, by covering diverse topics ranging from devices, channel behavior, communication and networking, to physical testbeds and demonstration systems. The key aspects presented in this paper will enable THz communications as a pillar of 6G and Beyond wireless systems in the next decade.

eess.SP↗

arXiv2020

Mobility-aware Beam Steering in Metasurface-based Programmable Wireless Environments

Programmable wireless environments (PWEs) utilize electromagnetic metasurfaces to transform wireless propagation into a software-controlled resource. In this work we study the effects of user device mobility on the efficiency of PWEs. An analytical model is proposed, which describes the potential misalignment between user-emitted waves and the active PWE configuration, and can constitute the basis for studying queuing problems in PWEs. Subsequently, a novel, beam steering approach is proposed which can effectively mitigate the misalignment effects. Ray-tracing-based simulations evaluate the proposed scheme.

cs.ET↗

arXiv2019

3D Channel Modeling and Characterization for Hypersurface Empowered Indoor Environment at 60 GHz Millimeter-Wave Band

This paper proposes a three-dimensional (3D) communication channel model for an indoor environment considering the effect of the Hypersurface. The Hypersurface is a software controlled intelligent metasurface, which can be used to manipulate electromagnetic waves, as for example for non-specular reflection and full absorption. Thus it can control the impinging rays from a transmitter towards a receiver location in both LOS and NLOS paths, e.g. to combat distance and improve wireless connectivity. We focus on the 60 GHz mmWave frequency band due to its increasing significance in 5G/6G networks and evaluate the effect of Hypersurface in an indoor environment in terms of attenuation coefficients related to the Hypersurface reflection and absorption functionalities, using CST simulation, a 3D electromagnetic simulator of high frequency components. To highlight the benefits of Hypersurface coated walls versus plain walls, we use the derived Hypersurface 3D channel model and a custom 3D ray-tracing simulator for plain walls considering a typical indoor scenario for different Tx-Rx location and separation distances.

eess.SP↗

arXiv2019

An Interpretable Neural Network for Configuring Programmable Wireless Environments

Software-defined metasurfaces (SDMs) comprise a dense topology of basic elements called meta-atoms, exerting the highest degree of control over surface currents among intelligent panel technologies. As such, they can transform impinging electromagnetic (EM) waves in complex ways, modifying their direction, power, frequency spectrum, polarity and phase. A well-defined software interface allows for applying such functionalities to waves and inter-networking SDMs, while abstracting the underlying physics. A network of SDMs deployed over objects within an area, such as a floorplan walls, creates programmable wireless environments (PWEs) with fully customizable propagation of waves within them. This work studies the use of machine learning for configuring such environments to the benefit of users within. The methodology consists of modeling wireless propagation as a custom, interpretable, back-propagating neural network, with SDM elements as nodes and their cross-interactions as links. Following a training period the network learns the propagation basics of SDMs and configures them to facilitate the communication of users within their vicinity.

cs.ET↗

arXiv2019

Intelligent Environments based on Ultra-Massive MIMO Platforms for Wireless Communication in Millimeter Wave and Terahertz Bands

Millimeter-wave (30-300 GHz) and Terahertz-band communications (0.3-10 THz) are envisioned as key wireless technologies to satisfy the demand for Terabit-per-second (Tbps) links in the 5G and beyond eras. The very large available bandwidth in this ultra-broadband frequency range comes at the cost of a very high propagation loss, which combined with the low power of mm-wave and THz-band transceivers limits the communication distance and data-rates. In this paper, the concept of intelligent communication environments enabled by Ultra-Massive MIMO platforms is proposed to increase the communication distance and data-rates at mm-wave and THz-band frequencies. An end-to-end physical model is developed by taking into account the capabilities of novel intelligent plasmonic antenna arrays which can operate in transmission, reception, reflection and waveguiding, as well as the peculiarities of the mm-wave and THz-band multi-path channel. Based on the developed model, extensive quantitative results for different scenarios are provided to illustrate the performance improvements in terms of both achievable distance and data-rate in Ultra-Massive MIMO environments.

eess.SP↗

arXiv2019

Combating the Distance Problem in the Millimeter Wave and Terahertz Frequency Bands

In the millimeter wave (30-300 GHz) and Terahertz (0.1-10 THz) frequency bands, high spreading loss and molecular absorption often limit the signal transmission distance and coverage range. In this paper, four directions to tackle the crucial problem of distance limitation are investigated, namely, a physical layer distance-aware design, ultra-massive MIMO communication, reflectarrays, and intelligent surfaces. Additionally, the potential joint design of these technologies is proposed to combine the benefits and possibly further extend the communication distance. Qualitative analyses and quantitative simulations are provided to illustrate the benefits of the proposed techniques and demonstrate the feasibility of mm-wave and THz band communications up to 100 meters in both line-of-sight and non-line-of-sight areas.

eess.SP↗