arXiv subjects

Chenyang Yang

Publications and source records attributed to Chenyang Yang.

At least 19 recordsLinked to original sources

arXiv2026

Cross-System Neural Precoder: Exploiting Structural Consistency for Fast Adaptation

Adapting learning-based precoding across different system configurations is challenging due to multiple types of variables and constraints. While large-scale neural networks have been proposed for cross-task adaptation, whether such adaptability requires large models remains unclear. In this paper, we identify a structural property of a class of precoding problems: the subproblems associated with each type of variable in alternative optimization (AO) share a common computational structure across systems when other variables are fixed. This structural consistency enables the reuse of update rules across systems. Based on this observation, we propose a cross-system neural precoder (XNP), where each layer implements AO-inspired update equations, which define the layer-wise input-output mappings. By reusing common update structures and learning only lightweight nonlinear mappings, the XNP enables efficient adaptation across systems only with several thousand trainable parameters. Simulation results show that pre-trained XNPs achieve fast adaptation to new configurations with significantly fewer training samples and epochs than a graph neural network-based baseline. This demonstrates that cross-system adaptability can be achieved by exploiting shared computational structure, rather than relying on large models.

eess.SP↗

arXiv2026

Positional Attention-based Graph Neural Network for Learning Permutation Non-equivariant Wireless Policies

Graph neural networks (GNNs) have emerged as a promising approach to learning wireless policies efficiently by leveraging topology prior and incorporating relational inductive biases. However, when the optimal policy is not permutation equivariant (PE), conventional GNNs suffer from mismatched inductive biases, leading to degraded performance or poor generalizability. This issue arises in wireless tasks with expected objectives, such as channel estimation and end-to-end (E2E) precoding, where the PE property of the optimal policy depends on the underlying channel distribution. In this paper, we propose a novel positional attention-based GNN to learn permutation nonequivariant policies efficiently. The core idea is to incorporate relative positions of vertices into the attention mechanism via an embedding function, enabling the GNNs to capture asymmetric relationships. Consequently, the proposed GNN can represent permutation non-equivariant functions, while retaining high learning efficiency and size generalizability through parameter sharing. We consider channel estimation and E2E precoding as case studies, and prove that their policies are PE to users but not to antennas under spatially correlated channels. We employ the proposed GNN to learn the policies, where the embedding function is designed based on the channel covariance matrix. Simulation results demonstrate that the proposed GNN outperforms existing channel estimation and E2E precoding methods, requires fewer samples for training, and can be generalized to systems with different numbers of antennas and users.

eess.SP↗

arXiv2026

Better Harnesses, Smaller Models: Building 90% Cheaper Agents via Automated Harness Adaptation

Frontier LLM agents are automating many business tasks, but their high inference cost makes large-scale deployment unsustainable. Small language models (SLMs) offer a cheaper alternative, yet they typically fall short when swapped into a harness designed for a frontier LLM. We show that for many routine business tasks, SLM agents can match LLM performance at 90% lower cost, when paired with an adapted harness that can be automatically discovered by a meta agent. The key insight is that much of the task difficulty is shared across instances and can be lifted from the model into the harness via tailored instructions, tools, and orchestration loops. To study this systematically, we create a framework that maps agent failure modes to harness adaptation strategies, and build a harness optimizer that automatically discovers effective adaptations from failure trajectories. Across seven business-oriented agentic tasks and three SLM families, we found optimized harnesses significantly improve performance on 16 of 21 task-SLM pairs, with seven pairs closing the SLM-LLM performance gap and the best SLM agent recovering 89.7% of LLM performance at 4% of the cost. Our analysis further shows that adaptation works best for tasks with more repetitive workflows and for SLMs with sufficient base capabilities. Together, these results suggest that harness adaptation can expand the practical deployment range of SLM agents in routine business tasks.

cs.SE↗

arXiv2026

Detecting AI-Generated Content on Social Media with Multi-modal Language Models

Generative AI has enabled the creation of photorealistic images and videos that are increasingly disseminated on social media, often used for spam, misinformation, manipulation, and fraud. Existing AI-generated content (AIGC) detection methods face challenges including poor generalization to new generation models, reliance on single modalities, and lack of interpretable explanations. We present our pipeline that mitigates these issues by continuously curating diverse multi-modal social media data and training a compact vision-language model for detection and explanation. Our model achieves state-of-the-art detection performance on public benchmarks and demonstrates robust detection and explanation capabilities on internal social media datasets across multiple platforms. We deployed our model for post recommendation on social media platforms and observed positive downstream impacts on user engagement, demonstrating that it is feasible to perform effective AIGC detection in dynamic, real-world social media environments.

cs.CL↗

arXiv2025

Joint Optimization of Pilot Length, Pilot Assignment, and Power Allocation for Cell-free MIMO Systems with Graph Neural Networks

In user-centric cell-free multi-antenna systems, pilot contamination degrades spectral efficiency (SE) severely. To mitigate pilot contamination, existing works jointly optimize pilot assignment and power allocation by assuming fixed pilot length, which fail to balance pilot overhead against the contamination. To maximize net-SE, we jointly optimize pilot length, pilot assignment, and power allocation with deep learning. Since the pilot length is a variable, the size of pilot assignment matrix is unknown during the optimization. To cope with the challenge, we design size-generalizable graph neural networks (GNNs). We prove that pilot assignment policy is a one-to-many mapping, and improperly designed GNNs cannot learn the optimal policy. We tackle this issue by introducing feature enhancement. To improve learning performance, we design a contamination-aware attention mechanism for the GNNs. Given that pilot assignment and power allocation respectively depend on large- and small-scale channels, we develop a dual-timescale GNN framework to explore the potential. To reduce inference time, a single-timescale GNN is also designed. Simulation results show that the designed GNNs outperform existing methods in terms of net-SE, training complexity, and inference time, and can be well generalized across problem scales and channels.

eess.SP↗

arXiv2025

"I Don't Think RAI Applies to My Model'' -- Engaging Non-champions with Sticky Stories for Responsible AI Work

Responsible AI (RAI) tools -- checklists, templates, and governance processes -- often engage RAI champions, individuals intrinsically motivated to advocate ethical practices, but fail to reach non-champions, who frequently dismiss them as bureaucratic tasks. To explore this gap, we shadowed meetings and interviewed data scientists at an organization, finding that practitioners perceived RAI as irrelevant to their work. Building on these insights and theoretical foundations, we derived design principles for engaging non-champions, and introduced sticky stories -- narratives of unexpected ML harms designed to be concrete, severe, surprising, diverse, and relevant, unlike widely circulated media to which practitioners are desensitized. Using a compound AI system, we generated and evaluated sticky stories through human and LLM assessments at scale, confirming they embodied the intended qualities. In a study with 29 practitioners, we found that, compared to regular stories, sticky stories significantly increased time spent on harm identification, broadened the range of harms recognized, and fostered deeper reflection.

cs.HC↗

arXiv2025

Modular PE-Structured Learning for Cross-Task Wireless Communications

Recent trends in learning wireless policies attempt to develop deep neural networks (DNNs) for handling multiple tasks with a single model. Existing approaches often rely on large models, which are hard to pre-train and fine-tune at the wireless edge. In this work, we challenge this paradigm by leveraging the structured knowledge of wireless problems -- specifically, permutation equivariant (PE) properties. We design three types of PE-aware modules, two of which are Transformer-style sub-layers. These modules can serve as building blocks to assemble compact DNNs applicable to the wireless policies with various PE properties. To guide the design, we analyze the hypothesis space associated with each PE property, and show that the PE-structured module assembly can boost the learning efficiency. Inspired by the reusability of the modules, we propose PE-MoFormer, a compositional DNN capable of learning a wide range of wireless policies -- including but not limited to precoding, coordinated beamforming, power allocation, and channel estimation -- with strong generalizability, low sample and space complexity. Simulations demonstrate that the proposed modular PE-based framework outperforms relevant large model in both learning efficiency and inference time, offering a new direction for structured cross-task learning for wireless communications.

eess.SP↗

arXiv2025

Understanding Prompt Programming Tasks and Questions

Prompting foundation models (FMs) like large language models (LLMs) have enabled new AI-powered software features (e.g., text summarization) that previously were only possible by fine-tuning FMs. Now, developers are embedding prompts in software, known as prompt programs. The process of prompt programming requires the developer to make many changes to their prompt. Yet, the questions developers ask to update their prompt is unknown, despite the answers to these questions affecting how developers plan their changes. With the growing number of research and commercial prompt programming tools, it is unclear whether prompt programmers' needs are being adequately addressed. We address these challenges by developing a taxonomy of 25 tasks prompt programmers do and 51 questions they ask, measuring the importance of each task and question. We interview 16 prompt programmers, observe 8 developers make prompt changes, and survey 50 developers. We then compare the taxonomy with 48 research and commercial tools. We find that prompt programming is not well-supported: all tasks are done manually, and 16 of the 51 questions -- including a majority of the most important ones -- remain unanswered. Based on this, we outline important opportunities for prompt programming tools.

cs.SE↗

arXiv2025

When Attention is Beneficial for Learning Wireless Resource Allocation Efficiently?

Owing to the use of attention mechanism to leverage the dependency across tokens, Transformers are efficient for natural language processing. By harnessing permutation properties broadly exist in resource allocation policies, each mapping measurable environmental parameters (e.g., channel matrix) to optimized variables (e.g., precoding matrix), graph neural networks (GNNs) are promising for learning these policies efficiently in terms of scalability and generalizability. To reap the benefits of both architectures, there is a recent trend of incorporating attention mechanism with GNNs for learning wireless policies. Nevertheless, is the attention mechanism really needed for resource allocation? In this paper, we strive to answer this question by analyzing the structures of functions defined on sets and numerical algorithms, given that the permutation properties of wireless policies are induced by the involved sets (say user set). In particular, we prove that the permutation equivariant functions on a single set can be recursively expressed by two types of functions: one involves attention, and the other does not. We proceed to re-express the numerical algorithms for optimizing several representative resource allocation problems in recursive forms. We find that when interference (say multi-user or inter-data stream interference) is not reflected in the measurable parameters of a policy, attention needs to be used to model the interference. With the insight, we establish a framework of designing GNNs by aligning with the structures. By taking reconfigurable intelligent surface-aided hybrid precoding as an example, the learning efficiency of the proposed GNN is validated via simulations.

eess.SP↗

arXiv2025

Adversarial Training: Enhancing Out-of-Distribution Generalization for Learning Wireless Resource Allocation

Unsupervised learning has been extensively adopted to train deep neural networks (DNNs) for learning wireless resource allocation. Yet, the performance of DNNs is vulnerable to distribution shifts between training and test data, e.g., wireless channels. In this paper, we propose an offline unsupervised training method to enhance the out-of-distribution (OOD) generalizability of DNNs. Inspired by adversarial training (AT), the method trains DNNs using progressively identified adversarial examples out of the training distribution. To reflect OOD degradation of a DNN in the context of unsupervised learning, we reformulate the optimization problem of AT. The proposed method is evaluated by learning hybrid precoding. Simulation results showcase the enhanced OOD performance of multiple kinds of DNNs with approximately 5\(\sim\)20\% improvement across various channel distributions, even when the samples only from a single distribution (e.g., Rayleigh fading) are used for training.

eess.SP↗

arXiv2025

cAST: Enhancing Code Retrieval-Augmented Generation with Structural Chunking via Abstract Syntax Tree

Retrieval-Augmented Generation (RAG) has become essential for large-scale code generation, grounding predictions in external code corpora to improve actuality. However, a critical yet underexplored aspect of RAG pipelines is chunking -- the process of dividing documents into retrievable units. Existing line-based chunking heuristics often break semantic structures, splitting functions or merging unrelated code, which can degrade generation quality. We propose chunking via Abstract Syntax Trees (\ourwork), a structure-aware method that recursively breaks large AST nodes into smaller chunks and merges sibling nodes while respecting size limits. This approach generates self-contained, semantically coherent units across programming languages and tasks, improving performance on diverse code generation tasks, e.g., boosting Recall@5 by 4.3 points on RepoEval retrieval and Pass@1 by 2.67 points on SWE-bench generation. Our work highlights the importance of structure-aware chunking for scaling retrieval-enhanced code intelligence.

cs.SE↗

arXiv2025

What Prompts Don't Say: Understanding and Managing Underspecification in LLM Prompts

Prompt underspecification is a common challenge when interacting with LLMs. In this paper, we present an in-depth analysis of this problem, showing that while LLMs can often infer unspecified requirements by default (41.1%), such behavior is fragile: Under-specified prompts are 2x as likely to regress across model or prompt changes, sometimes with accuracy drops exceeding 20%. This instability makes it difficult to reliably build LLM applications. Moreover, simply specifying all requirements does not consistently help, as models have limited instruction-following ability and requirements can conflict. Standard prompt optimizers likewise provide little benefit. To address these issues, we propose requirements-aware prompt optimization mechanisms that improve performance by 4.8% on average over baselines. We further advocate for a systematic process of proactive requirements discovery, evaluation, and monitoring to better manage prompt underspecification in practice.

cs.CL↗

arXiv2025

SPHERE: An Evaluation Card for Human-AI Systems

In the era of Large Language Models (LLMs), establishing effective evaluation methods and standards for diverse human-AI interaction systems is increasingly challenging. To encourage more transparent documentation and facilitate discussion on human-AI system evaluation design options, we present an evaluation card SPHERE, which encompasses five key dimensions: 1) What is being evaluated?; 2) How is the evaluation conducted?; 3) Who is participating in the evaluation?; 4) When is evaluation conducted?; 5) How is evaluation validated? We conduct a review of 39 human-AI systems using SPHERE, outlining current evaluation practices and areas for improvement. We provide three recommendations for improving the validity and rigor of evaluation practices.

cs.HC↗

arXiv2025

Precoder Learning by Leveraging Unitary Equivariance Property

Incorporating mathematical properties of a wireless policy to be learned into the design of deep neural networks (DNNs) is effective for enhancing learning efficiency. Multi-user precoding policy in multi-antenna system, which is the mapping from channel matrix to precoding matrix, possesses a permutation equivariance property, which has been harnessed to design the parameter sharing structure of the weight matrix of DNNs. In this paper, we study a stronger property than permutation equivariance, namely unitary equivariance, for precoder learning. We first show that a DNN with unitary equivariance designed by further introducing parameter sharing into a permutation equivariant DNN is unable to learn the optimal precoder. We proceed to develop a novel non-linear weighting process satisfying unitary equivariance and then construct a joint unitary and permutation equivariant DNN. Simulation results demonstrate that the proposed DNN not only outperforms existing learning methods in learning performance and generalizability but also reduces training complexity.

eess.SP↗

arXiv2025

Optimizing QoE-Privacy Tradeoff for Proactive VR Streaming

Proactive virtual reality (VR) streaming requires users to upload viewpoint-related information, raising significant privacy concerns. Existing strategies preserve privacy by introducing errors to viewpoints, which, however, compromises the quality of experience (QoE) of users. In this paper, we first delve into the analysis of the viewpoint leakage probability achieved by existing privacy-preserving approaches. We determine the optimal distribution of viewpoint errors that minimizes the viewpoint leakage probability. Our analyses show that existing approaches cannot fully eliminate viewpoint leakage. Then, we propose a novel privacy-preserving approach that introduces noise to uploaded viewpoint prediction errors, which can ensure zero viewpoint leakage probability. Given the proposed approach, the tradeoff between privacy preservation and QoE is optimized to minimize the QoE loss while satisfying the privacy requirement. Simulation results validate our analysis results and demonstrate that the proposed approach offers a promising solution for balancing privacy and QoE.

cs.MM↗

arXiv2025

Quantization Design for Deep Learning-Based CSI Feedback

Deep learning-based autoencoders have been employed to compress and reconstruct channel state information (CSI) in frequency-division duplex systems. Practical implementations require judicious quantization of encoder outputs for digital transmission. In this paper, we propose a novel quantization module with bit allocation among encoder outputs and develop a method for joint training the module and the autoencoder. To enhance learning performance, we design a loss function that adaptively weights the quantization loss and the logarithm of reconstruction loss. Simulation results show the performance gain of the proposed method over existing baselines.

eess.SP↗

arXiv2025

Distributed Resource Block Allocation for Wideband Cell-free System

This paper studies distributed resource block (RB) allocation in wideband orthogonal frequency-division multiplexing (OFDM) cell-free systems. We propose a novel distributed sequential algorithm and its two variants, which optimize RB allocation based on the information obtained through over-the-air (OTA) transmissions between access points (APs) and user equipments, enabling local decision updates at each AP. To reduce the overhead of OTA transmission, we further develop a distributed deep learning (DL)-based method to learn the RB allocation policy. Simulation results demonstrate that the proposed distributed algorithms perform close to the centralized algorithm, while the DL-based method outperforms existing baseline methods.

eess.SP↗

arXiv2025

Learning of Uplink Resource Allocation with Multiuser QoS Constraints

In the paper the joint optimization of uplink multiuser power and resource block (RB) allocation are studied, where each user has quality of service (QoS) constraints on both long- and short-blocklength transmissions. The objective is to minimize the consumption of RBs for meeting the QoS requirements, leading to a mixed-integer nonlinear programming (MINLP) problem. We resort to deep learning to solve the problem with low inference complexity. To provide a performance benchmark for learning based methods, we propose a hierarchical algorithm to find the global optimal solution in the single-user scenario, which is then extended to the multiuser scenario. The design of the learning method, however, is challenging due to the discrete policy to be learned, which results in either vanishing or exploding gradient during neural network training. We introduce two types of smoothing functions to approximate the involved discretizing processes and propose a smoothing parameter adaption method. Another critical challenge lies in guaranteeing the QoS constraints. To address it, we design a nonlinear function to intensify the penalties for minor constraint violations. Simulation results demonstrate the advantages of the proposed method in reducing the number of occupied RBs and satisfying QoS constraints reliably.

eess.SP↗