Showing 1–50 of 749 results for author: Wong, K

Automated Estimation of MBIST Area and Test Time in Heterogeneous Memory IPs via Stacked Ensemble Framework

Authors: Chee Jin Teoh, Ab Al-Hadi Ab Rahman, Johnny Kee Hui Wong, Premkumar A/L Kesavan Prabagaran, Muhammad Nadzir Marsono, Nuzhat Khan

Abstract: Embedded memories occupy a large portion of modern System-on-Chip (SoC) designs, especially in high-performance applications such as artificial intelligence and edge computing. Memory Built-In Self-Test (MBIST) is commonly used to ensure memory reliability, but it introduces additional area and test time overhead. Accurate early estimation of these overheads is important during design planning, ye… ▽ More Embedded memories occupy a large portion of modern System-on-Chip (SoC) designs, especially in high-performance applications such as artificial intelligence and edge computing. Memory Built-In Self-Test (MBIST) is commonly used to ensure memory reliability, but it introduces additional area and test time overhead. Accurate early estimation of these overheads is important during design planning, yet conventional methods rely on full Register Transfer Level (RTL) synthesis and test pattern generation, which are slow and resource-intensive. This study proposes a supervised learning framework that predicts MBIST area and test time directly from RTL-level design parameters without synthesis. A dataset of 4,470 samples for area and 624 for test time was generated using Synopsys Design Compiler and MINT, an Intel-enhanced MBIST tool. Input features include memory count, word width, address depth, port configuration, and clock domains. For area prediction, the features are processed through polynomial expansion, log transformation, and scaling, followed by a stacked ensemble model using XGBoost, LightGBM, and a Neural Network with Gradient Boosting as the meta-learner. For test time, XGBoost and LightGBM are combined using Ridge Regression, with hyperparameters tuned through a 100-trial Optuna search. The models achieved 90.68% accuracy for area and 96.80% for test time within a +/-10% margin, improving over baseline methods by 8.53% and 48.80% respectively. The results show that this approach enables faster estimation of MBIST costs and supports more efficient design decisions in memory IP development. △ Less

Submitted 20 August, 2026; originally announced August 2026.

Write, Execute, Refine: From Skill Followers to Skill Optimizers via Reinforcement Learning from Execution Feedback

Authors: Kang Peng, Zhiwei Zhang, Yichen Zhang, Zezhong Wang, Yiming Du, Geng Tu, Baojun Wang, Bin Liang, Ruifeng Xu, Kam-Fai Wong

Abstract: Expert-written natural language skills can improve tool-using agents, yet agent-authored skills perform 8-11 points worse than using no skill. This gap suggests that following procedural guidance and improving it from execution evidence are distinct capabilities. Inference time loops can repair skills but do not improve the model that writes the next one. We study how to organize execution experie… ▽ More Expert-written natural language skills can improve tool-using agents, yet agent-authored skills perform 8-11 points worse than using no skill. This gap suggests that following procedural guidance and improving it from execution evidence are distinct capabilities. Inference time loops can repair skills but do not improve the model that writes the next one. We study how to organize execution experience from intermediate skills into training states for an optimizer. We introduce WER (Write, Execute, and Refine), a multi-phase framework that trains a Skill Optimizer outside a frozen executor. The optimizer proposes skills, a frozen agent executes each repeatedly, and a programmatic verifier scores the outcomes. The scores provide relative credit and select mixed-outcome records. Matched successful and failed trajectories from these records form the next phase's refinement states, so the optimizer learns from the consequences of its earlier outputs. On BFCL v4 multi-turn and tau2-bench, WER improves average Pass@1 over the no-skill baseline by 7.80 and 3.85 points, respectively. Under an identical refinement workflow, it outperforms the same backbone without optimizer training by 9.35 and 10.29 points. The trained 4B optimizer reaches 76.63 percent on BFCL v4, outperforming all evaluated off-the-shelf general-purpose models used as skill optimizers on average. △ Less

Submitted 18 August, 2026; originally announced August 2026.

LEGO-RL: Harness-Native Reinforcement Learning for Coding Agents

Authors: Yiming Du, Yuxin Jiang, Tao Yuan, Jianbo Dai, Shaowei Wang, Jierun Chen, Chaofan Tao, Xianzhi Yu, Lifeng Shang, Kam-Fai Wong, Xiaohui Li, Haoli Bai

Abstract: Reinforcement learning for coding agents increasingly relies on long-running agent harnesses to manage tool integration, repository contexts, and execution feedback. However, the native execution environments of these harnesses are inherently misaligned with policy-gradient training: environmental crashes and reward hacking corrupt outcome signals, while train-inference discrepancies decouple roll… ▽ More Reinforcement learning for coding agents increasingly relies on long-running agent harnesses to manage tool integration, repository contexts, and execution feedback. However, the native execution environments of these harnesses are inherently misaligned with policy-gradient training: environmental crashes and reward hacking corrupt outcome signals, while train-inference discrepancies decouple rollout behavior from policy updates. To address this, we present LEGO-RL, a framework that bridges native coding-agent harnesses with scalable policy-gradient optimization without modifying their internal control flow. LEGO-RL is built upon three pillars: (1) faithful optimization via in-process LLM proxying that captures raw generation streams for token-level alignment and robust trainer-side log-probability recomputation, even under harness-side compaction or re-serialization; (2) reliable execution via scalable sandbox orchestration featuring image caching and stage-wise defenses to mitigate reward hacking; and (3) observable training through an integrated plugin that automates validation and monitoring, paired with a Live UI for granular trajectory diagnostics. We evaluate LEGO-RL by training the sparse MoE model Qwen3.5-35B-A3B with GSPO across three native coding-agent harnesses. LEGO-RL improves Qwen3.5-35B-A3B across OpenHands SDK (64.0% to 70.4%), Claude Code (62.4% to 68.2%), and OpenCode (57.2% to 66.6%) on SWE-bench Verified, while maintaining a rollout-training probability correlation above 0.99. △ Less

Submitted 18 August, 2026; originally announced August 2026.

Comments: Webpage: https://lego-rl.pages.dev

Mitigating Context Interference for Reliable and Efficient Search Agents

Authors: Boyang Xue, Bin Wu, Shuofei Qiao, Sheng Wang, Rui Wang, Yiming Du, Hongru Wang, Jeff Z. Pan, Emine Yilmaz, Kam-Fai Wong, Aldo Lipani

Abstract: Recent research empowers Large Language Models (LLMs) as multi-turn search agents to iteratively retrieve and generate outputs until complex tasks are solved. However, the contexts of multi-turn search agents are lengthy and complex. For example, the retrieved set of documents in each turn would inevitably introduce irrelevant information that distracts LLMs, referring to \textit{context interfere… ▽ More Recent research empowers Large Language Models (LLMs) as multi-turn search agents to iteratively retrieve and generate outputs until complex tasks are solved. However, the contexts of multi-turn search agents are lengthy and complex. For example, the retrieved set of documents in each turn would inevitably introduce irrelevant information that distracts LLMs, referring to \textit{context interference}, potentially hindering the reliability and efficiency of search agents. Therefore, we conduct a systematic study on context interference in multi-turn search agents, focusing on investigating i) which parts of the context of search agents will contribute to the context interference, ii) how to refine the contexts of search agents to mitigate the interference, and iii) can incorporating context refinement into search agent training yield further improvements. We reveal that interference primarily arises from the latest retrieved documents. Based on the explored findings, we then introduce a distill-based context refiner to dynamically mitigate context interference for multi-turn search agents. Finally, we validate that incorporating context refinement into RL training pipelines of search agents can significantly enhance both reliability and efficiency. This study highlights the importance of mitigating context interference of search agents, inspiring a novel paradigm of ``refine context and then generate'' for AI agents. △ Less

Submitted 11 August, 2026; originally announced August 2026.

CSI Reconstruction in Fluid Antenna Systems Without Spatial Covariance Priors

Authors: Zhentian Zhang, Kaitao Meng, Tuo Wu, Kai-Kit Wong, Hao Xu, Liang Liu, Pei Xiao, Chao Wang, Kin-Fai Tong

Abstract: Fluid antenna systems (FASs) exploit many candidate ports for spatial diversity, but hardware constraints allow channel observations at only a few active ports. Whether full-port CSI can be recovered without pre-acquired channel statistics remains open. Under the Clarke isotropic scattering model, we show that the channel lies in a low-dimensional spatial modal subspace determined by the scatterin… ▽ More Fluid antenna systems (FASs) exploit many candidate ports for spatial diversity, but hardware constraints allow channel observations at only a few active ports. Whether full-port CSI can be recovered without pre-acquired channel statistics remains open. Under the Clarke isotropic scattering model, we show that the channel lies in a low-dimensional spatial modal subspace determined by the scattering environment rather than the total port count. Consequently, recovery becomes feasible when the number of observed ports reaches the modal dimension (i.e., $M\geq r$), even when $M\ll N$. We further establish a sharp feasibility threshold: reliable recovery is impossible below this dimension regardless of SNR, whereas accuracy improves with additional observations above it. By decomposing the recovery error into modal truncation, estimation, and learning components, we derive explicit tradeoffs among RF chains, pilot overhead, transmit power, and training data. These results enable scalable prior-free full-port CSI recovery with few active ports. △ Less

Submitted 10 August, 2026; originally announced August 2026.

Modeling and Performance Analysis for Fluid Antenna System Enabled UAV Near-Field Communications

Authors: Hao Jiang, Wangqi Shi, Zhentian Zhang, Xusheng Zhu, Kai-Kit Wong, Hyundung Shin

Abstract: Fluid antenna systems (FASs) offer a promising solution for unmanned aerial vehicle (UAV) air-to-ground (A2G) communications by enabling reconfigurable radiation characteristics. Addressing the limitations of traditional models in capturing the dynamic port configuration of FAS and the near-field nature of UAV communications, this paper proposes a dynamic port-reconfigurable near-field channel mod… ▽ More Fluid antenna systems (FASs) offer a promising solution for unmanned aerial vehicle (UAV) air-to-ground (A2G) communications by enabling reconfigurable radiation characteristics. Addressing the limitations of traditional models in capturing the dynamic port configuration of FAS and the near-field nature of UAV communications, this paper proposes a dynamic port-reconfigurable near-field channel model for FAS-assisted UAV-to-mobile user (MU) links. Furthermore, we develop a FAS-adaptive subarray partition scheme utilizing a greedy strategy. By decomposing line-of-sight (LoS) and non-line-of-sight (NLoS) components and integrating UAV motion dynamics with FAS port activation states, the proposed model accurately characterizes the non-uniform spatial distribution of near-field channels. The subarray partition scheme dynamically groups active ports to satisfy near-field conditions while significantly reducing computational complexity, supported by a dynamic update algorithm that efficiently handles subarray adjustments during port switching. To avoid low effective gain and deep-fading ports in dense FAS configurations, a channel gain-based selection strategy is employed to prioritize high-gain ports. We derive and analyze the modeling accuracy and channel capacity, investigating the impact of FAS dimensions, port spacing, active port count, and UAV dynamics on system performance. Finally, the computational complexity of the subarray partition scheme is evaluated, verifying its advantages for real-time applications and providing a theoretical foundation for the design and analysis of FAS in dynamic scenarios. △ Less

Submitted 10 August, 2026; originally announced August 2026.

SoRoMoX: Fast, Differentiable, and Parallelizable Soft Robot Models

Authors: Maximilian Stölzle, Solange Gribonval, Daniel Feliu-Talegon, Vito Daniele Perfetta, Michele Martini, Chuhan Zhang, Kiwan Wong, Mohammed Tarnini, Anup Teejo Mathew, Federico Renda, Daniela Rus, Cosimo Della Santina

Abstract: Reduced-order models based on Cosserat-rod theory are now well established, and modeling theory is no longer the primary bottleneck in soft-robot control. Their implementations, however, do not support the differentiable, GPU-parallel, and control-oriented workflows that underpin advanced rigid-robotics applications. Here, we fill this gap with SoRoMoX (Soft Robot Models in JAX), a fully numerical… ▽ More Reduced-order models based on Cosserat-rod theory are now well established, and modeling theory is no longer the primary bottleneck in soft-robot control. Their implementations, however, do not support the differentiable, GPU-parallel, and control-oriented workflows that underpin advanced rigid-robotics applications. Here, we fill this gap with SoRoMoX (Soft Robot Models in JAX), a fully numerical, JIT-compilable Python/JAX framework. SoRoMoX implements articulated, Piecewise Constant Strain, and Variable Strain models through a unified, control-ready interface that provides inertia matrices, gravitational and elastic forces, Jacobians, and their derivatives. To our knowledge, it is the first rod/strain-based soft-robot modeling framework that runs directly on GPUs and is end-to-end differentiable with respect to states, inputs, and parameters. Sequential CPU rollouts are up to 18.1x faster than state-of-the-art alternatives, while GPU-parallel rollouts increase throughput by up to 234.6x. This performance enables workflows that were previously impractical or impossible: static-equilibrium system identification with 66% lower marker RMSE; residual-force learning with a further 64% reduction; computed-torque tracking with RMSE reduced by a factor of approximately 500 relative to model-free PD; control-gain optimization with up to 62% lower loss than untuned gains; safety-constrained control using high-order control barrier functions to keep the peak contact force within a prescribed 5 N bound, compared with 33.5 N without the safety constraint; and reinforcement-learning policy training up to 7x faster than a CPU PyElastica discrete-rod baseline through massively parallel rollouts. △ Less

Submitted 6 August, 2026; originally announced August 2026.

SyncPlan: Long-Horizon LLM Coordination with Explicit Synchronization and Adaptive Correction

Authors: Shen You, Xiaoming Zhu, Weining Weng, Hefei Mei, Weixuan Wang, Zhongshen Li, Zeji LI, Ye-Wen Wang, Zijun Liao, Juchao Zhuo, Yang Wei, Fuhao Qiu, Siqin Li, Zhenjie Lian, Danei Gong, Junkai Ji, Xiangtao Li, Qiuzhen Lin, Liang Wang, Ka-Chun Wong

Abstract: LLM-based multi-agent coordination faces a fundamental trade-off between efficiency and adaptivity in dynamic environments. Existing approaches typically rely on repeated LLM invocations or multi-round communication to adapt decisions during execution, introducing substantial latency and making coordination vulnerable to asynchronous progress and environmental changes. Conversely, one-shot plannin… ▽ More LLM-based multi-agent coordination faces a fundamental trade-off between efficiency and adaptivity in dynamic environments. Existing approaches typically rely on repeated LLM invocations or multi-round communication to adapt decisions during execution, introducing substantial latency and making coordination vulnerable to asynchronous progress and environmental changes. Conversely, one-shot planning reduces coordination overhead but produces open-loop plans that can quickly become stale or fail when actions depend on other agents and the environment. We introduce SyncPlan, a plan-execute-correct framework for long-horizon coordination through explicit synchronization and adaptive correction. Given the state and team-level task, a centralized LLM coordinator generates per-agent action chains in a single planning call. During execution, explicit wait primitives and deadlock detection enforce inter-agent and agent-environment dependencies, while a lightweight Plan Staleness Detector continuously assesses the remaining plan and triggers replanning when environmental changes invalidate its assumptions. We further optimize the coordinator through SFT and planning-oriented RL with dense task progress and outcome-level execution feedback. Experiments on the public Overcooked benchmark and the complex Honor of Kings environment show that SyncPlan achieves state-of-the-art task success rates while using less than 0.05% of the wall-clock runtime compared with existing LLM-based coordinators. Code and datasets will be made publicly available. △ Less

Submitted 2 August, 2026; originally announced August 2026.

Physically Real-time Infrared Attack against Optical Flow Estimation Networks

Authors: Shen You, Wei Jiang, Jiarui Liu, Yijian Ye, Qiuzhen Lin, Xiangtao Li, Ka-Chun Wong

Abstract: With the promising performance of deep neural networks on image-based tasks, different real-world applications such as autonomous driving and motion detection have become increasingly mature and relevant to human lives. In particular, Optical Flow Estimation Networks (OFENs), as upstream models, play a critical role in different domains. Its outputs are heavily assumed and adopted for different do… ▽ More With the promising performance of deep neural networks on image-based tasks, different real-world applications such as autonomous driving and motion detection have become increasingly mature and relevant to human lives. In particular, Optical Flow Estimation Networks (OFENs), as upstream models, play a critical role in different domains. Its outputs are heavily assumed and adopted for different downstream tasks, and it is essential to test its robustness to prevent safety accidents. We present an approach for real-time attacks on OFENs in the physical world, leveraging infrared lights for their stealthiness. By generating a large number of Adversarial Examples in advance, our approach computes AEs in real time and dynamically displays them, which allows our method to facilitate precise and targeted attacks without modifying the victim system. Unlike previous digital-to-physical attack techniques, our method directly attacks victim models within the physical world, thereby overcoming the limitations associated with the ineffectiveness of AEs. Experimental results demonstrate the efficacy of our approach in compromising OFENs across diverse lighting conditions, varying object motion velocities, and different object placements, ultimately impairing the network's ability to accurately estimate optical flow. △ Less

Submitted 29 July, 2026; originally announced July 2026.

MultivationBench: A Benchmark for Multimodal Sequential Motivation Reasoning

Authors: Kawai Chung, Chunkit Chan, Yauwai Yim, Yuxuan Liu, Haochen Shi, Weiqi Wang, Qing Zong, Tianshi Zheng, Yixuan Fu, Kai Chung Wong, Hao Liang, Yifan Gao, Xi Yang, Janet Hui-wen Hsiao, Yangqiu Song

Abstract: Multimodal Large Language Models have sparked significant interest due to their potential for social intelligence; however, their ability to perform sequential motivation reasoning remains insufficiently studied. Existing evaluations predominantly examine static text or isolated visual snapshots, which do not reflect the cumulative nature of real-world behavioral drivers. To address this gap, we i… ▽ More Multimodal Large Language Models have sparked significant interest due to their potential for social intelligence; however, their ability to perform sequential motivation reasoning remains insufficiently studied. Existing evaluations predominantly examine static text or isolated visual snapshots, which do not reflect the cumulative nature of real-world behavioral drivers. To address this gap, we introduce MultivationBench, a benchmark designed to rigorously evaluate multimodal motivation reasoning within story-driven visual narratives. The benchmark builds upon established psychological frameworks - Maslow's hierarchy and Reiss's basic desires - and requires models to integrate accumulated multimodal context to infer evolving motivations. Results indicate that MultivationBench presents a significant challenge: all tested models struggle to maintain consistent motivation reasoning across sequential contexts, revealing a critical disconnect between static recognition capabilities and the dynamic reasoning essential for human-like social understanding. △ Less

Submitted 29 July, 2026; originally announced July 2026.

Comments: 31 pages, including appendices; 6 figures and 22 tables. Code: https://github.com/HKUST-KnowComp/MultivationBench

Learned Blockwise Port Activation for Real Time Beamforming in Fluid Antenna Arrays

Authors: Yuanhui Wu, Zhentian Zhang, Hanjiang Hong, Hao Jiang, Zaichen Zhang, Kai-Kit Wong, Yin Xu, Wenjun Zhang

Abstract: Fluid antenna arrays (FAAs), support multiuser downlink transmission by activating a subset of reconfigurable ports. The activation mask jointly determines the effective channel and the sparse radiating aperture, which requires a balance among sum rate, sidelobe suppression, hardware constraints, and online complexity. Channel driven selection can cluster active ports and increase sidelobes, where… ▽ More Fluid antenna arrays (FAAs), support multiuser downlink transmission by activating a subset of reconfigurable ports. The activation mask jointly determines the effective channel and the sparse radiating aperture, which requires a balance among sum rate, sidelobe suppression, hardware constraints, and online complexity. Channel driven selection can cluster active ports and increase sidelobes, whereas sidelobe oriented synthesis is typically channel independent and can sacrifice sum rate. This paper proposes learned blockwise port activation (L-BPA), for real time sidelobe aware FAA downlink beamforming. L-BPA activates a fixed number of ports in each aperture block, which supports grouped switching hardware and limits port clustering. A lightweight convolutional network scores ports using multiuser channel features, port coordinates, and user power statistics. Training combines blockwise straight through masks with a differentiable peak sidelobe level (PSLL), surrogate. During inference, learned scores are combined with multiscale geometric repulsion, followed by regularized zero forcing precoding over the reduced effective channel. L-BPA reduces the average PSLL by 3.26 dB relative to uniform sparse activation while achieving a slightly higher sum rate. It also reduces the PSLL by 8.13 dB and 10.10 dB relative to greedy and gain based selection, respectively, without iterative online search. △ Less

Submitted 28 July, 2026; originally announced July 2026.

Expectation Alignment of Language Models for Real-World User Expectations

Authors: Miaomiao Li, Yang Wang, Bin Liang, Shudong Liu, Zhiwei Zhang, Kam-Fai Wong

Abstract: Large language models (LLMs) have demonstrated remarkable performance on standard benchmarks, yet it remains largely unexplored whether they truly meet user expectations. Existing evaluation approaches, relying on model heuristics, expert rubrics, or user simulation, fail to capture the diversity and subtlety of real human expectations, causing models to appear competent while misaligning with wha… ▽ More Large language models (LLMs) have demonstrated remarkable performance on standard benchmarks, yet it remains largely unexplored whether they truly meet user expectations. Existing evaluation approaches, relying on model heuristics, expert rubrics, or user simulation, fail to capture the diversity and subtlety of real human expectations, causing models to appear competent while misaligning with what users actually seek. We present the first systematic study of user expectations in real-world LLM interactions, proposing a principled procedure to extract semantically rich expectations and introducing ExpectBench, a benchmark grounded in real user expectations. Analyses reveal that current LLMs struggle to satisfy and anticipate what users hope to obtain, highlighting a fundamental source of misalignment. Building on these observations, we propose LENS, a lightweight latent expectation-aware response generation framework. LENS enables models to internalize user expectations and generate better-aligned responses, consistently improving expectation satisfaction and underscoring the importance of explicitly modeling user expectations for realistic human-AI alignment. △ Less

Submitted 2 June, 2026; originally announced July 2026.

Comments: Accepted by ICML 2026

On the Performance of Fluid Antenna Systems under Block-Diagonal Correlation Model

Authors: Jiangsheng Huangfu, Zhengyu Song, Tianwei Hou, Anna Li, Kai Kit Wong

Abstract: Fluid antenna systems (FASs) have recently emerged as a promising reconfigurable antenna technology for future wireless networks, owing to their unique ability to exploit fine-grained spatial channel variations within a compact aperture. In this paper, a single-input multiple-output (SIMO) FAS employing maximum-ratio combining (MRC) is investigated under the block-diagonal correlation model, where… ▽ More Fluid antenna systems (FASs) have recently emerged as a promising reconfigurable antenna technology for future wireless networks, owing to their unique ability to exploit fine-grained spatial channel variations within a compact aperture. In this paper, a single-input multiple-output (SIMO) FAS employing maximum-ratio combining (MRC) is investigated under the block-diagonal correlation model, where the ports of FAS are partitioned into independent blocks and the strongest port within each block is selected for MRC combining. Exact outage probability (OP) expressions are first derived in both convolution and characteristic-function forms. To gain further insights, closed-form high-SNR asymptotic expressions are developed, from which the diversity order is shown to approximate the number of ports. This result reveals that block partitioning influences only the coding gain and can therefore be optimized without compromising the diversity performance. For the ergodic rate (ER), a Gamma-matching approximation together with a tighter Jensen-based approximation is derived in closed form. Simulation results corroborate the analytical framework and demonstrate that: i) increasing either the number of ports or the number of blocks improves the system performance; ii) the diversity order depends solely on the number of ports; and iii) the proposed SIMO-FAS achieves comparable or superior outage performance to conventional MRC receivers despite employing fewer combining branches. △ Less

Submitted 21 July, 2026; originally announced July 2026.

HARQ for Slow Fluid Antenna Multiple Access

Authors: Sixu Han, Kai-Kit Wong, Hanjiang Hong, Hyundong Shin

Abstract: Slow fluid antenna multiple access (sFAMA), enabled by the fluid antenna system (FAS), has recently emerged as a practical and low-complexity paradigm for supporting massive wireless connectivity. While existing studies have characterized its physical-layer performance under one-shot transmission, its interaction with retransmission protocols and the resulting networking performance remain largely… ▽ More Slow fluid antenna multiple access (sFAMA), enabled by the fluid antenna system (FAS), has recently emerged as a practical and low-complexity paradigm for supporting massive wireless connectivity. While existing studies have characterized its physical-layer performance under one-shot transmission, its interaction with retransmission protocols and the resulting networking performance remain largely unexplored. In this paper, we study a downlink hybrid automatic repeat request (HARQ)-assisted sFAMA framework, termed HARQ-sFAMA, in which each user performs distinguished port selection in every HARQ round and combines the received signals across multiple rounds to improve decoding reliability. We develop a comprehensive analytical framework to characterize the outage probability, average packet waiting time, and energy efficiency of the proposed system. The analysis reveals how HARQ exploits the spatial reconfigurability of FAS to simultaneously enhance reliability and improve queueing performance. Numerical results corroborate the theoretical analysis and demonstrate that the HARQ-sFAMA system significantly outperforms conventional one-shot sFAMA in terms of reliability, delay, and energy efficiency. These findings suggest that the integration of HARQ and sFAMA provides a promising pathway toward a practical and standards-compatible massive access solution for future wireless networks. △ Less

Submitted 20 July, 2026; originally announced July 2026.

RRAM-DP: Device-Calibrated Differential Privacy for In-Memory Edge Learning

Authors: Kwunhang Wong, Jichang Yang, Karl M. H. Lai, Hegan Chen, Songqi Wang, Wei Xuan, Ning Lin, Han Wang, Xiaojuan Qi, Zhongrui Wang

Abstract: Edge Artificial Intelligence of Things (AIoT) systems often collect sensitive data in situ, raising serious privacy concerns. Resistive-switching random-access memory (RRAM) is an attractive substrate for efficient AIoT thanks to its multi-bit storage and compute-in-memory (CiM) capabilities, while its inherently stochastic write behavior provides a natural source of randomness that can be leverag… ▽ More Edge Artificial Intelligence of Things (AIoT) systems often collect sensitive data in situ, raising serious privacy concerns. Resistive-switching random-access memory (RRAM) is an attractive substrate for efficient AIoT thanks to its multi-bit storage and compute-in-memory (CiM) capabilities, while its inherently stochastic write behavior provides a natural source of randomness that can be leveraged for differential privacy (DP) protection. Yet how to transform this device-level randomness-typically viewed as detrimental to accuracy-into a principled randomized mechanism while preserving model utility remains underexplored. We propose RRAM-DP, a hardware-algorithm co-design that relaxes RRAM write-verify operations to inject calibrated noise for inherently (epsilon, delta)-DP with formal DP analysis; together with pretraining techniques, it renders a novel private, high-utility CiM training paradigm. On CIFAR-10/100, STS-B, and SST-2, RRAM-DP-SGD incurs at best only a 3.8% accuracy drop at (epsilon=2, delta=O(1/n))-DP relative to non-private SGD. At the same privacy level, RRAM-DP-SGD delivers up to 57x and 3.2x energy savings and 2.7x and 1.8x speedups over A100 and DiVa-GEMM, respectively. These results point toward efficient, privacy-preserving in-memory training on RRAM at the edge. △ Less

Submitted 31 July, 2026; v1 submitted 20 July, 2026; originally announced July 2026.

Comments: International Conference on Computer-Aided Design 2026

Detection, Attribution, Narration: An End-to-End Pipeline for Explainable Money Mule Identification

Authors: Yuge Zhang, Yuanxing Zhang, Yichao Jin, Khairul Amsyar Mohd Razis, Nicholas Qi An Choo, Kai Yin Anders Wong, Xinyan Tang, Kenneth Zhu Ke, Wee Keong Dennis Lee, Jingyuan Zhao

Abstract: Money mule accounts are critical facilitators of financial fraud, yet detecting them at scale remains challenging due to the heterogeneous nature of transactional and behavioural data. We present an end-to-end pipeline for customer-level mule detection comprising three stages: (1) a LightGBM classifier trained on 280 engineered features spanning transaction patterns, account demographics, network… ▽ More Money mule accounts are critical facilitators of financial fraud, yet detecting them at scale remains challenging due to the heterogeneous nature of transactional and behavioural data. We present an end-to-end pipeline for customer-level mule detection comprising three stages: (1) a LightGBM classifier trained on 280 engineered features spanning transaction patterns, account demographics, network topology, and temporal behaviour; (2) a TreeSHAP attribution layer that decomposes each prediction into feature contributions; and (3) a large language model (LLM) module that converts SHAP attributions into analyst-facing natural-language narratives. We evaluate across three open-weight LLM families and assess explanation quality through analyst feedback. In a live production deployment, the system achieves a yield rate of 89%, up from 61% under the incumbent rule-based system, with monthly alert volume expanding from 211 to 302, reflecting broader true-positive coverage rather than increased noise. This corresponds to a 60% incremental adverse detection beyond existing review workflows, substantially outperforming the rule-based approach. Qualitative feedback from analysts indicates that LLM-generated narratives reduce cognitive load during alert triage. We further discuss implications of deploying LLM-augmented explainability in regulated financial environments. △ Less

Submitted 20 July, 2026; originally announced July 2026.

Beyond Implicit Force: Evaluating Explicit Force-Torque Proxies in Action Chunking with Transformers

Authors: King Hang Wong, Lingqiao Liu, Feras Dayoub

Abstract: Contact-rich manipulation requires policies to infer interaction state from signals that are often weakly observable through vision and kinematics alone. Action Chunking with Transformers (ACT) has shown strong performance in fine-grained manipulation, but many deployments collect demonstrations through leader-follower teleoperation, where tracking error between commanded leader motion and execute… ▽ More Contact-rich manipulation requires policies to infer interaction state from signals that are often weakly observable through vision and kinematics alone. Action Chunking with Transformers (ACT) has shown strong performance in fine-grained manipulation, but many deployments collect demonstrations through leader-follower teleoperation, where tracking error between commanded leader motion and executed follower motion implicitly encodes contact, resistance, and constraint violation. This paper examines whether ACT's apparent force-awareness depends on this hidden interaction cue. We introduce an observation-centric ACT variant that predicts future follower joint states instead of leader commands, thereby removing the teleoperation-induced discrepancy signal while preserving the rest of the learning pipeline. We then evaluate whether simple joint-torque proxies, derived from onboard motor current or joint effort, can recover contact-aware behavior without external force/torque sensors. Across four real-world tasks spanning surface following, insertion, stiffness discrimination, and force-based stopping, removing the implicit cue leads to severe failures in force-critical phases. In contrast, torque-augmented policies recover robust contact behavior and improve the base ACT policy. These results demonstrate that, on real hardware, the implicit teleoperation cue is a recoverable source of force-awareness, where torque signals are available, a simple proxy matches, surpasses, or further enhances it. △ Less

Submitted 16 July, 2026; originally announced July 2026.

Comments: Accepted to IROS 2026

Reinforcing Egocentric Spatial Perception in Multimodal Large Language Models via Ego Scene Augmentation

Authors: Chi Kit Wong, Ye Pan, Yuanhuiyi Lyu, Xu Zheng, Zidong Cao, Lutao Jiang, Zixin Zhang, Huiyu Zhou, Xuming Hu

Abstract: Egocentric Visual Question Answering (VQA) has attracted widespread attention as an important task for enabling Multimodal Large Language Models (MLLMs) to interact with the real world. However, existing MLLMs struggle to perform effective spatial reasoning in complex egocentric scenes due to their limited spatial perception capabilities. To this end, we introduce Ego Scene Augmentation (ESA), an… ▽ More Egocentric Visual Question Answering (VQA) has attracted widespread attention as an important task for enabling Multimodal Large Language Models (MLLMs) to interact with the real world. However, existing MLLMs struggle to perform effective spatial reasoning in complex egocentric scenes due to their limited spatial perception capabilities. To this end, we introduce Ego Scene Augmentation (ESA), an egocentric spatial perception framework, which actively enhances the spatial perception capabilities from the egocentric perspective, powered by the proposed Ego-element Graph. Our core insight is leveraging the Ego-element Graph as an intermediary representation to augment the egocentric spatial perception of MLLMs via visual foundational models. Specifically, we 1) construct the Ego-element Graph, which encapsulates and integrates egocentric spatial features enabled by visual foundational models; 2) enhance the spatial perception capabilities of MLLMs via the Ego-element Graph for ego-perspective scenes. Our proposed ESA framework presents significant performance improvement on the EgoTextVQA benchmark. We achieve an 8.14% gain on the indoor setting and an 8.72% gain on the outdoor setting. Furthermore, our ESA shows the most impressive performance improvement in the shopping subset of the indoor setting. The project code is publicly available. △ Less

Submitted 20 August, 2026; v1 submitted 15 July, 2026; originally announced July 2026.

Comments: 14 pages, 8 figures. Chi Kit Wong and Ye Pan contributed equally. Code: https://github.com/Chikit-WONG/spatialGraph

Demystifying On-Policy Distillation: Roles, Pathologies, and Regulations

Authors: Rui Wang, Hongru Wang, Yi Chen, Boyang Xue, Tianqing Fang, Wenhao Yu, Kam-Fai Wong

Abstract: On-policy distillation (OPD) has become a key paradigm in LLM post-training, yet its training dynamics remain poorly understood. We present a systematic study examining the role, pathologies, and regulations of OPD. We first clarify the role of OPD as an exploration catalyst: it steers the student toward correct reasoning paths via dense token-level guidance, without expanding capability ceiling.… ▽ More On-policy distillation (OPD) has become a key paradigm in LLM post-training, yet its training dynamics remain poorly understood. We present a systematic study examining the role, pathologies, and regulations of OPD. We first clarify the role of OPD as an exploration catalyst: it steers the student toward correct reasoning paths via dense token-level guidance, without expanding capability ceiling. We confirm this by showing that prompt diversity matters more than per-problem sampling numbers, and critically, that the effectiveness of OPD hinges entirely on the quality of its guiding signal. This dependency exposes two pathologies that derail exploration. The Student-Teacher Mismatch occurs when a large teacher-student distributional gap causes the guiding signal to misalign with task correctness, steering exploration in counterproductive directions. Length Exploitation arises when the aggregated token-level objective creates length-dependent shortcuts, allowing the student to game the reward landscape through response truncation or redundant padding, exploring degenerate length modes rather than reasoning strategies. To tame these pathologies, we investigate lightweight signal regulations: advantage clipping and log-scale compression, ensuring exploration is guided by faithful signals. Experiments across seven benchmarks demonstrate that these regulations alleviate length exploitation and enable effective distillation, stably surpassing OPD variants and RLVR baselines, thereby confirming that well-regulated signal quality, rather than mere teacher scale, governs successful exploration in OPD. △ Less

Submitted 14 July, 2026; originally announced July 2026.

HERO: A Heterogeneity-Aware Benchmark Library for Federated Continual Learning

Authors: Thinh T. H. Nguyen, Le-Tuan Nguyen, Minh-Duong Nguyen, Nhi Trinh, Anh Tran Nam Nguyet, Dung D. Le, Kok-Seng Wong

Abstract: Federated continual learning (FCL) evaluates how distributed clients learn from changing data streams while retaining previously learned knowledge. Existing evaluations are difficult to compare because they often change datasets, task splits, client data splits, task orders, backbones, memory assumptions, and reporting rules simultaneously. We introduce \textbf{HERO}, a heterogeneity-aware benchma… ▽ More Federated continual learning (FCL) evaluates how distributed clients learn from changing data streams while retaining previously learned knowledge. Existing evaluations are difficult to compare because they often change datasets, task splits, client data splits, task orders, backbones, memory assumptions, and reporting rules simultaneously. We introduce \textbf{HERO}, a heterogeneity-aware benchmark library for FCL. HERO builds benchmark streams by separating three choices that are often coupled, namely the task split, the client data split, and the client task sequence. In HERO-Core, the main comparable benchmark, $α$ controls client data skew and $ρ$ controls task-order mismatch. We evaluate representative FCL methods on CIFAR-100 and TinyImageNet using final average accuracy, average forgetting, and bottom-10\% client accuracy. We also include a graph-based Domain-IL portability case study on OGB-MolPCBA, where scaffold-domain granularity changes the input distribution while the prediction task remains fixed. Our results show that method behavior changes across easy and heterogeneous settings, that average accuracy can hide weak bottom-client performance, that task-order mismatch favors different strategies from synchronized evaluation, and that the same HERO interface can expose domain-shift difficulty beyond image-based FCIL. HERO releases benchmark streams, configurations, method implementations, and reporting scripts to support reproducible and setting-aware FCL evaluation. △ Less

Submitted 13 June, 2026; originally announced July 2026.

Comments: 30 pages, 10 figures

Learning Subset-Shared Invariances for Domain Generalization with Mixture-of-Experts

Authors: Tien-Hung Nguyen, Tien-Dat Tran, M. -Duong Nguyen, Kok-Seng Wong

Abstract: Domain generalization (DG) aims to learn a model from one or more source domains that generalizes to an unseen target domain without accessing target data during training. A common approach enforces invariance of representations across all source domains, assuming predictive structure is globally shared. However, we demonstrate that enforcing invariance across more domains gradually restricts the… ▽ More Domain generalization (DG) aims to learn a model from one or more source domains that generalizes to an unseen target domain without accessing target data during training. A common approach enforces invariance of representations across all source domains, assuming predictive structure is globally shared. However, we demonstrate that enforcing invariance across more domains gradually restricts the feasible representation space, discarding transferable predictive factors that are not universally shared. To address this limitation, we propose subset-shared invariance, where predictive structure is assumed stable only within domain subsets. We implement this principle with a mixture-of-experts architecture, where each expert aligns the specific domains it serves and a routing mechanism composes subset-invariant components for prediction. This creates a routing-conditioned invariance, jointly learned with the representation. To facilitate effective decomposition, we develop training objectives that encourage selective alignment, confident and balanced routing, and diverse expert specialization. Experiments on DomainBed benchmarks demonstrate improved out-of-domain generalization and greater robustness under increasing domain heterogeneity. Our results suggest that DG should move beyond enforcing a single global invariance and instead model invariance through partially shared structure across domain subsets. △ Less

Submitted 24 June, 2026; originally announced June 2026.

Rethinking Molecular Graph Backdoors under Chemistry-aware Admission

Authors: Thinh T. H. Nguyen, Sze Jue Yang, Khoa D. Doan, Chee Seng Chan, Kok-Seng Wong

Abstract: Backdoor attacks on molecular graph neural networks (GNNs) are typically evaluated as abstract graph edits, but real molecular learning pipelines do not train on arbitrary graphs. Molecular records must first survive parsing, sanitization, canonicalization, and graph-string consistency checks. We formalize this overlooked admission stage as ChemGuard, an operational protocol for testing whether a… ▽ More Backdoor attacks on molecular graph neural networks (GNNs) are typically evaluated as abstract graph edits, but real molecular learning pipelines do not train on arbitrary graphs. Molecular records must first survive parsing, sanitization, canonicalization, and graph-string consistency checks. We formalize this overlooked admission stage as ChemGuard, an operational protocol for testing whether a submitted molecular record can enter a realistic learning pipeline, while complementing existing defenses. ChemGuard admits a record only when its molecular string is sanitizable and the graph reconstructed from that string matches the submitted molecular graph. Under this operational view, many existing graph-based backdoors lose much of their apparent efficacy because their poisons are chemically invalid or representation-inconsistent. We then show that admission checks alone are insufficient to rule out molecular backdoors. We propose ChemBack, an admission-aware molecular backdoor attack that constructs chemically feasible motif-anchor attachments and ranks admitted candidates by fingerprint-based Tanimoto similarity to clean target-class molecules. ChemBack is model-free during trigger selection, using molecular structures, target labels, fingerprints, and public validity checks, but no victim model, surrogate GNN, learned embedding, gradient, logit, or training-code access. Across molecular benchmarks, validators, architectures, and defenses, \textbf{ChemBack} achieves high attack success with fully admitted poisons while preserving clean accuracy. Our results reveal a two-sided lesson, chemistry-aware admission suppresses many graph-only backdoors, yet chemically valid and target-aligned molecular backdoors remain a practical threat. △ Less

Submitted 22 June, 2026; originally announced June 2026.

Comments: 30 pages

MSC Class: 68T05; 68T07; 68T10 ACM Class: I.2.6; I.5.1; K.6.5; J.2

Enormous Fluid Antenna Systems (E-FAS) for Wireless Sensing: Channel Modeling and Conditional Estimation Limits

Authors: Farshad Rostami Ghadi, Kai-Kit Wong, Jose D. Vega-Sanchez, Kin-Fai Tong, Hyundong Shin

Abstract: In this paper, we develop a fundamental analytical framework for integrated sensing and communications (ISAC) enabled by the Enormous Fluid Antenna System (E-FAS), which transforms a collection of coordinated intelligent surfaces into a gigantic reconfigurable electromagnetic aperture, with particular emphasis on the limits of angular sensing.We begin by developing a bidirectional sensing channel… ▽ More In this paper, we develop a fundamental analytical framework for integrated sensing and communications (ISAC) enabled by the Enormous Fluid Antenna System (E-FAS), which transforms a collection of coordinated intelligent surfaces into a gigantic reconfigurable electromagnetic aperture, with particular emphasis on the limits of angular sensing.We begin by developing a bidirectional sensing channel model that explicitly captures the complete sensing process, including surface-wave (SW) routing, distributed reradiation, target scattering, and echo propagation. Based on this channel model, we formulate a parametric observation model for target sensing and derive the associated Fisher information matrix (FIM) and Cramer-Rao bound (CRB) for angular estimation. The analysis demonstrates that E-FAS gives rise to a fundamentally different sensing regime compared with conventional array-based and reconfigurable-surface-aided ISAC architectures. Our analysis uncovers that maximizing coherent routing gain does not necessarily maximize sensing performance, exposing a fundamental trade-off between SW routing gain and sensing diversity in programmable propagation environments. Numerical results validate the developed framework and demonstrate that E-FAS-enabled ISAC systems can achieve substantial angular sensing gains over conventional architectures under the same transmit-power budget. The results further underscore the importance of jointly optimizing propagation routing and sensing functionality, positioning E-FAS as a new paradigm for ISAC. △ Less

Submitted 22 June, 2026; originally announced June 2026.

Spatial Modulation for Tx-SIMO-FAS: Port Selection and Performance Analysis

Authors: Xusheng Zhu, Kai-Kit Wong, Hanjiang Hong, Chenguang Rao, Kaitao Meng

Abstract: This paper considers a single-input multiple-output (SIMO) setup with a fluid antenna system (FAS) at the transmitter side and multiple fixed antennas at the receiver, which is referred to as a Tx-SIMO-FAS. We investigate the use of spatial modulation (SM) utilizing the FAS on a single radio-frequency (RF) chain while the receiver side performs maximum-likelihood detection. Unlike conventional ant… ▽ More This paper considers a single-input multiple-output (SIMO) setup with a fluid antenna system (FAS) at the transmitter side and multiple fixed antennas at the receiver, which is referred to as a Tx-SIMO-FAS. We investigate the use of spatial modulation (SM) utilizing the FAS on a single radio-frequency (RF) chain while the receiver side performs maximum-likelihood detection. Unlike conventional antenna arrays, however, the large number of fluid antenna ports accommodated within a limited aperture introduces strong spatial correlation, which reduces the distinguishability of port indices and degrades the reliability of index detection. To address this challenge, three correlation-aware port-selection schemes are proposed: successive fluid Euclidean-distance-optimized selection (SF-EDAS), successive orthogonal port selection (SOPS), and correlation-constrained orthogonal array selection (CC-COAS). These schemes focus on enhancing received-constellation separation, improving channel-basis conditioning, and jointly optimizing channel gain and inter-port decorrelation, respectively. To understand the performance limits of FAS-SM, a reliability analysis is developed by decomposing the channel into an energy-based degree of freedom (DoF), and an extreme-value DoF. High signal-to-noise ratio (SNR) analysis reveals an effective diversity order determined by the number of selected ports, the number of receive antennas, and the energy-based spatial DoF. Furthermore, the aperture-limited array gain is characterized through a scalar equivalent independent-look approximation involving the Digamma function. Numerical results demonstrate that the proposed schemes significantly outperform conventional SM and grouping-based benchmarks. Among them, CC-COAS achieves the most favorable tradeoff between error performance and computational complexity. △ Less

Submitted 28 June, 2026; v1 submitted 20 June, 2026; originally announced June 2026.

SEED: Simple ViT and Evolving Harness for Explainable Text Forgery Detection

Authors: Kahim Wong, Kemou Li, Yiming Chen, Haiwei Wu, Jiantao Zhou

Abstract: AI-assisted image editing threatens trust in financial, legal, and identity records. The GenText-Forensics Challenge at ACM MM 2026 addresses this by requiring structured forensic reports, in which integrating detection, pixel-level localization, and natural language explanation for multilingual text-centric forgery images. We present SEED, a modular system with three components. First, a similari… ▽ More AI-assisted image editing threatens trust in financial, legal, and identity records. The GenText-Forensics Challenge at ACM MM 2026 addresses this by requiring structured forensic reports, in which integrating detection, pixel-level localization, and natural language explanation for multilingual text-centric forgery images. We present SEED, a modular system with three components. First, a similarity-guided pipeline augments training with diverse synthetic forgeries. Second, a single ViT, built on DINOv3 with LoRA adaptation, jointly performs detection and pixel-level localization while preserving pre-trained priors with minimal trainable parameters. Third, an evolving harness takes the detector's predictions and generates a complete forensic report via an MLLM, iteratively improved through a proposer-evaluator loop optimizing report quality. SEED ranked 3rd in the GenText-Forensics Challenge. Code and data are available at https://github.com/KahimWong/GenText-Forensics-3rd-Place. △ Less

Submitted 5 August, 2026; v1 submitted 19 June, 2026; originally announced June 2026.

When Generator Replay Degrades: Projected Rehearsal Orchestration for Heterogeneous Federated Class-Incremental Learning

Authors: Thinh T. H. Nguyen, Khoa D. Doan, Binh T. Nguyen, Danh Le-Phuoc, Kok-Seng Wong

Abstract: Federated class-incremental learning (FCIL) becomes substantially harder when clients observe different label subsets, progress through tasks at different stages, and provide uneven supervision for the same semantic concepts. Existing FCIL methods often preserve old knowledge through input-space synthesis, but they can be fragile under heterogeneous task streams and difficult to transfer across mo… ▽ More Federated class-incremental learning (FCIL) becomes substantially harder when clients observe different label subsets, progress through tasks at different stages, and provide uneven supervision for the same semantic concepts. Existing FCIL methods often preserve old knowledge through input-space synthesis, but they can be fragile under heterogeneous task streams and difficult to transfer across modalities. To alleviate such issues, we propose PRO, a framework that replaces synthetic input replay with projected rehearsal orchestration. To remove external pretraining, we evaluate all methods under the same warmup. After this, PRO maintains compact class-level projected memories on the server and allows clients perform balanced pseudo multi-task training over current examples and old projected memories. To handle stronger representation drift, we further introduce PRO-MAX, which augments PRO with neighborhood-weighted memory alignment while preserving the same server-light principle that the server only aggregates model updates and memory statistics. Across image, text, and graph benchmarks, PRO and PRO-MAX improve retention and final utility under heterogeneous streams while remaining competitive in homogeneous FCIL. Even when baselines are given expanded replay budgets, they degrade under supervision imbalance and stage misalignment, indicating that replay quantity alone does not resolve replay-quality failures. Additional weak-task diagnostics further show that larger replay mismatch is associated with larger downstream degradation, while our method keeps projected memories better aligned with the evolving representation. △ Less

Submitted 14 June, 2026; originally announced June 2026.

Comments: 46 pages

MSC Class: 68T05 (Primary); 68T07; 68M14 (Secondary) ACM Class: I.2.6; I.2.11; C.2.4

Enhanced Fluid Index Modulation for Integrated Data and Energy Transfer

Authors: Long Zhang, Yizhe Zhao, Halvin Yang, Qiang Liu, Kai-Kit Wong

Abstract: Integrated data and energy transfer (IDET) is a promising technique for supporting sustainable low-power wireless networks. To improve both communication reliability and energy transfer efficiency, this paper investigates a fluid index modulation (FIM) assisted IDET system, where the base station employs a two-dimensional fluid antenna system (FAS) and the receiver adopts a power-splitting archite… ▽ More Integrated data and energy transfer (IDET) is a promising technique for supporting sustainable low-power wireless networks. To improve both communication reliability and energy transfer efficiency, this paper investigates a fluid index modulation (FIM) assisted IDET system, where the base station employs a two-dimensional fluid antenna system (FAS) and the receiver adopts a power-splitting architecture. In FIM, the information bits are delivered not only from the modulation symbols, but also the index of antenna position. Under finite-alphabet signaling, the average harvested power, bit error rate (BER), and achievable data rate are derived in closed form. A joint optimization problem is formulated to maximize the average harvested power subject to BER and achievable rate constraints by jointly optimizing the port selection, precoding vector, and power splitting ratio. An alternating optimization framework is developed, where the precoding vector and port selection are obtained via a Riemannian augmented Lagrangian method (RALM) and block coordinate descent (BCD) algorithm, respectively. Simulation results demonstrate that the proposed scheme achieves a superior rate-energy trade-off over benchmark schemes, while the proposed algorithm attains near-optimal performance with significantly lower complexity than exhaustive search. △ Less

Submitted 3 June, 2026; originally announced June 2026.

Low-rank Distributional Matrix Completion

Authors: Jiayi Wang, Raymond K. W. Wong

Abstract: We study a distributional generalization of the matrix completion problem in which each entry of the target matrix is a probability distribution rather than a scalar. In this setting, only a subset of matrix entries is observed, and even for observed entries, the underlying distributions are not directly accessible; instead, we observe finitely many samples drawn from them. To represent distributi… ▽ More We study a distributional generalization of the matrix completion problem in which each entry of the target matrix is a probability distribution rather than a scalar. In this setting, only a subset of matrix entries is observed, and even for observed entries, the underlying distributions are not directly accessible; instead, we observe finitely many samples drawn from them. To represent distributional entries, we employ kernel mean embeddings and introduce a notion of Tucker rank for distribution-valued matrices to capture their low-rank structure. The infinite-dimensional nature of kernel embeddings poses significant methodological challenges. To address this, we introduce functional unfolding operators that link the proposed distributional low-rank structure to the classical Tucker rank for finite-dimensional tensors. Based on this framework, we propose a novel estimator for distributional matrix completion. We establish non-asymptotic error bounds that characterize the statistical performance of the estimator. Extensive experiments on synthetic data and a real-world application demonstrate the effectiveness of the proposed method. △ Less

Submitted 2 June, 2026; originally announced June 2026.

Geometry-Structured Channel Reconstruction for Conventional and Fluid Antenna Systems: Bayesian Inference and Fundamental Limits

Authors: Zhentian Zhang, Kai-Kit Wong, Kaitao Meng, David Morales-Jimenez, Hao Jiang, Christos Masouros, Hyundong Shin, Zaichen Zhang

Abstract: Accurate channel state information (CSI) acquisition is critical for exploiting the spatial flexibility of fluid antenna systems (FASs). However, port selection and transmission optimization require CSI over a large number of candidate port positions, making direct port-wise estimation prohibitively costly in terms of pilot overhead. This paper addresses this challenge through geometry-structured… ▽ More Accurate channel state information (CSI) acquisition is critical for exploiting the spatial flexibility of fluid antenna systems (FASs). However, port selection and transmission optimization require CSI over a large number of candidate port positions, making direct port-wise estimation prohibitively costly in terms of pilot overhead. This paper addresses this challenge through geometry-structured channel reconstruction, which exploits the fact that the port-domain CSI can be parameterized by a small number of dominant propagation paths. We first establish fundamental mean square error (MSE) and normalized MSE (NMSE) benchmarks for both geometry-structured and unstructured channel reconstruction, providing analytical references for evaluating the intrinsic benefit of geometric modeling in conventional antenna systems and FASs. Motivated by the strong spatial correlation induced by densely distributed fluid antenna ports, we further propose a Bayesian reconstruction framework, termed geometry-structured expectation-maximization approximate message passing (GS-EM-AMP). The proposed algorithm incorporates geometric channel structure into the EM-AMP procedure and adaptively learns unknown statistical parameters from noisy observations. Numerical results demonstrate that GS-EM-AMP achieves near-bound reconstruction accuracy while maintaining strong robustness against steering-domain correlation, thereby offering an efficient and reliable solution for large-scale CSI acquisition in FASs. △ Less

Submitted 26 May, 2026; originally announced June 2026.

Bridging Auxiliary Constraints to Resolve Instruction Following in Large Reasoning Models

Authors: Zhengyi Zhao, Shubo Zhang, Huimin Wang, Zezhong Wang, Yutian Zhao, Yefeng Zheng, Binyang Li, Yulan He, Kam-Fai Wong, Xian Wu

Abstract: Large Reasoning Models (LRMs) have demonstrated impressive capabilities in many tasks, yet they struggle with reliably following multiple instructions, either by failing to satisfy individual constraints or by struggling to balance competing constraints simultaneously. We formalize this challenge as the Constraint Adherence Problem (CAP). This paper introduces a novel framework that addresses CAP… ▽ More Large Reasoning Models (LRMs) have demonstrated impressive capabilities in many tasks, yet they struggle with reliably following multiple instructions, either by failing to satisfy individual constraints or by struggling to balance competing constraints simultaneously. We formalize this challenge as the Constraint Adherence Problem (CAP). This paper introduces a novel framework that addresses CAP by representing instructions as a structured knowledge graph of constraints. Our approach, Constraint Relationship Graph Completion (CRGC), explicitly models relationships between constraints, identifies adherence challenges, and discovers ``bridge constraints'' that help the model better focus on and reconcile requirements. Bridge constraints act as auxiliary instructions that make primary constraints more salient and compatible. Unlike existing approaches that enhance instruction following through general training methods, CRGC specifically improves constraint satisfaction by leveraging the model's own knowledge to create better pathways for generation. Experiments across three popular instruction following datasets demonstrate that our approach reduces constraint violations by 39% compared to standard prompting while maintaining reasoning abilities of large reasoning models. △ Less

Submitted 2 June, 2026; originally announced June 2026.

Comments: a pre-MIT Press publication version

Beyond the Literal: Decomposing Pragmatic Intent in Multimodal Meme Understanding

Authors: Zhengyi Zhao, Shubo Zhang, Zezhong Wang, Luyao Ye, Huimin Wang, Hanqi Yan, Binyang Li, Kam-Fai Wong, Yulan He

Abstract: When asked what a meme or sarcastic post means, Large Vision Language Models (LVLMs) tend to describe what the image shows rather than what the author is trying to communicate. Standard instruction tuning entangles a post's literal content with its pragmatic meaning, letting surface-level details contaminate the final response. We reframe meme understanding as a problem of literal-pragmatic decomp… ▽ More When asked what a meme or sarcastic post means, Large Vision Language Models (LVLMs) tend to describe what the image shows rather than what the author is trying to communicate. Standard instruction tuning entangles a post's literal content with its pragmatic meaning, letting surface-level details contaminate the final response. We reframe meme understanding as a problem of literal-pragmatic decomposition and propose \textbf{Intent Projection}, a framework that separates the two signals at the representation, output, and objective levels within a single LVLM backbone. At the representation level, an orthogonal projection module removes dominant unimodal directions from the fused image-text representation, retaining only the pragmatic residual, while a surface-real affect classifier anchors the decoder with a discrete tag that names the polarity gap. At the output level, the model externalizes a structured reasoning chain, and at the objective level a contrastive reward explicitly penalizes answers that restate the literal description. Across six multimodal benchmarks, Intent Projection consistently outperforms open-source baselines and narrows the gap to proprietary models, with the largest gains on high-divergence posts where literal collapse is most damaging. △ Less

Submitted 2 June, 2026; originally announced June 2026.

Revisiting Neural Processes via Fourier Transform and Volterra Series

Authors: Peiman Mohseni, Nick Duffield, Raymond K. W. Wong

Abstract: Modeling unknown latent functions from finite, irregularly sampled measurements is a recurring challenge across science and engineering. Neural processes (NPs), a family of probabilistic functional models, are promising solutions -- especially when endowed with domain-specific symmetries like translation equivariance, which improve sample efficiency and generalization. Yet existing translation-equ… ▽ More Modeling unknown latent functions from finite, irregularly sampled measurements is a recurring challenge across science and engineering. Neural processes (NPs), a family of probabilistic functional models, are promising solutions -- especially when endowed with domain-specific symmetries like translation equivariance, which improve sample efficiency and generalization. Yet existing translation-equivariant NPs face two limitations: (i) they stack generic components with non-linearities, obscuring the induced function class and limiting interpretability; and (ii) convolutional designs are limited by local receptive fields and the need to embed inputs onto a dense uniform grid, while attention-based alternatives lift these restrictions at quadratic cost in the number of observations. We address both with two contributions. First, using the Volterra expansion, we approximate continuous translation-equivariant operators by sums of higher-order convolutions, yielding analytical transparency while admitting efficient evaluation via first-order convolutions. Second, we introduce set Fourier convolutions (SFConvs), a frequency-domain parameterization that operates directly on irregularly sampled points, achieves approximately global receptive fields, and scales linearly in the number of observations. Building on these ideas, we propose two conditional NPs (CNPs): SFConvCNPs, which stack SFConv blocks with non-linearities, and SFVConvCNPs, which integrate the Volterra formulation. Experiments on synthetic and real-world datasets demonstrate our methods' efficacy against state-of-the-art baselines. △ Less

Submitted 13 July, 2026; v1 submitted 31 May, 2026; originally announced June 2026.

DecMem: Towards Minute-Long Consistent World Generation with Decoupled Memory

Authors: Zhenhao Yang, Xiaoshi Wu, Zhengyao Lv, Xiaoyu Shi, Xintao Wang, Pengfei Wan, Kun Gai, Kwan-Yee K. Wong

Abstract: Recent advances in video generative models have promoted rapid progress in controllable world models. However, maintaining fine-grained spatio-temporal consistency under long-horizon reasoning remains a key challenge. In this work, we move beyond explicit 3D memory and coarse frame-level implicit modeling, and propose a fine-grained, learnable, and scalable memory for consistent world generation.… ▽ More Recent advances in video generative models have promoted rapid progress in controllable world models. However, maintaining fine-grained spatio-temporal consistency under long-horizon reasoning remains a key challenge. In this work, we move beyond explicit 3D memory and coarse frame-level implicit modeling, and propose a fine-grained, learnable, and scalable memory for consistent world generation. We first identify two fundamental limitations of naïve learnable memory architectures in long-horizon extrapolation, namely computational inefficiency and attention dispersion. Through a systematic analysis of attention dispersion, we propose DecMem, a decoupled memory architecture that employs Sparse Global Memory for efficient fine-grained access to global history and Anchored Local Memory for stable and high-quality extrapolation. Extensive experiments demonstrate that DecMem significantly outperforms current state-of-the-art methods. By ensuring precise and efficient long-term memory and achieving superior extrapolation capabilities, DecMem enables minute-level controllable long video generation with high fidelity and consistency. △ Less

Submitted 29 May, 2026; originally announced May 2026.

Comments: Project page is available at https://jeffreyyzh.github.io/DecMem-Page

AMix-2: Establishing Protein as a Native Modality in Large Language Models

Authors: Keyue Qiu, Yixin Wu, Lihao Wang, Yawen Ouyang, Jixiang Yu, Zihan Zhou, Changze Lv, Dongyu Xue, Yuxuan Song, Xinbo Zhang, Hao Wang, Jiangtao Feng, Zhiqiang Gao, Lijun Wu, Xiaoqing Zheng, Ka-Chun Wong, Lei Bai, Ya-Qin Zhang, Wei-Ying Ma, Dahua Lin, Bowen Zhou, Hao Zhou

Abstract: We present AMix-2, a protein-text foundation model that establishes protein as a native modality in large language models (LLMs), unifying protein understanding and sequence design within a single foundation model. AMix-2 is built upon two key ideas: (1) a unified protein-text formulation that embeds natural language and protein sequence in a shared token space, enabling one model to perform biolo… ▽ More We present AMix-2, a protein-text foundation model that establishes protein as a native modality in large language models (LLMs), unifying protein understanding and sequence design within a single foundation model. AMix-2 is built upon two key ideas: (1) a unified protein-text formulation that embeds natural language and protein sequence in a shared token space, enabling one model to perform biological reasoning and conditional design instead of separate downstream task-specialized models; and (2) a block-wise diffusion language modeling backbone that combines causal generation across blocks with bidirectional context and iterative refinement within blocks. This scheme better matches the intrinsic nature of proteins than a strict left-to-right factorization. To evaluate protein foundation models under realistic generalization settings, we further introduce ProteinArena, a comprehensive benchmark with time-aware and homology-aware protocols across various understanding and design tasks, and with baselines covering classical bioinformatics tools, protein-specialized models and LLMs. On ProteinArena, AMix-2 outperforms frontier LLMs and demonstrates competitive performance to task-specific protein models. Controlled experiments further show that the diffusion-based paradigm generally surpasses its autoregressive counterpart, highlighting the advantage of flexible generation order for protein sequences. We release both AMix-2 and ProteinArena to facilitate open research in protein foundation models. △ Less

Submitted 29 May, 2026; originally announced May 2026.

Comments: 30 pages, 4 figures, 12 tables

Density-aware Sample-specific Attack

Authors: Qiyuan Wang, Yao Li, Raymond K. W. Wong

Abstract: Despite recent progress in backdoor attacks, existing methods remain susceptible to post-training defenses that erase the backdoor through fine-tuning or pruning. We revisit the core objectives of backdoor attacks and derive principled criteria characterizing optimal sample-specific trigger construction under a Bayes-optimal model of the victim's training. Our analysis reveals that both attack suc… ▽ More Despite recent progress in backdoor attacks, existing methods remain susceptible to post-training defenses that erase the backdoor through fine-tuning or pruning. We revisit the core objectives of backdoor attacks and derive principled criteria characterizing optimal sample-specific trigger construction under a Bayes-optimal model of the victim's training. Our analysis reveals that both attack success and clean-accuracy preservation are simultaneously optimized when triggered samples are steered into low-density regions of the clean data distribution, a distributional condition that controls all moments of the poisoned distribution at once rather than a handful of input-space summary statistics. We introduce a bilevel optimization framework that estimates density ratios via conditional time-score matching and optimizes a mixture-model objective to place triggered samples in these sparse regions. Extensive evaluations on MNIST, CIFAR-10, GTSRB, and TinyImageNet demonstrate that our method achieves above 99\% attack success rate before defense and retains 50--85 percentage points higher post-defense ASR than the strongest baselines under fine-tuning defenses. Against neuron-pruning defenses, the method exhibits complete immunity, with zero neurons identified for removal across all pruning thresholds. These results expose a fundamental gap in current defense paradigms and underscore the need for defenses that operate beyond the support of the clean distribution. △ Less

Submitted 28 May, 2026; v1 submitted 26 May, 2026; originally announced May 2026.

ACM Class: I.2.6; K.6.5

PRISM: A Multi-Dimensional Benchmark for Evaluating LLM Peer Reviewers

Authors: Ngoc Phan Phuoc Loc, Toan Huynh La Viet, Thanh Tran Khanh, Duy A Nguyen, Tuan Anh Nguyen Pham, Thanh Nguyen, Nitesh V. Chawla, Wray Buntine, Kok-Seng Wong, Khoa D. Doan, Binh T. Nguyen

Abstract: The rapid growth in submissions to machine learning venues has strained the scientific peer-review system and intensified interest in LLM-based automated peer reviewers. However, how good these systems are actually, especially compared to human reviewers at catching scientific gaps, remains poorly understood. In this work, we introduce PRISM (Peer Review Intelligence via Structured Multi-dimension… ▽ More The rapid growth in submissions to machine learning venues has strained the scientific peer-review system and intensified interest in LLM-based automated peer reviewers. However, how good these systems are actually, especially compared to human reviewers at catching scientific gaps, remains poorly understood. In this work, we introduce PRISM (Peer Review Intelligence via Structured Multi-dimensional assessment), a benchmarking framework that evaluates review quality across four dimensions: Depth of Analysis, Novelty Assessment,Flaw Identification & Major Issues Prioritization, and Multi-dimensional Constructiveness. Unlike most existing evaluations based on surface-level metrics like ROUGE and BLEU, or unconstrained LLM-as-a-judge prompting that conflates fluency with rigor, PRISM grounds each dimension in argument mining, retrieval-augmented verification, and consensus-based scoring. We apply PRISM to benchmark five leading automated reviewer systems and human reviewers on a stratified corpus of reviews from ICLR, ICML, and NeurIPS. The results reveal that LLMs can match or beat human reviewers on individual dimensions: comparable depth of analysis, stronger novelty verification, and highly accurate critique prioritization. However, no single system consistently matches the balanced performance of the human baseline across all dimensions at once. Each exhibits a distinct specialization profile with characteristic blind spots -- failure modes that aggregate metrics miss entirely. The implication is that LLM reviewers are best understood as targeted supplements to human review, effective within specific dimensions, but unreliable as standalone replacements. Our demo and key results can be found at https://khanhthanhdev.github.io/prism-page/. △ Less

Submitted 27 May, 2026; v1 submitted 26 May, 2026; originally announced May 2026.

Brain-to-Image Retrieval and Reconstruction via Multimodal EEG Alignment

Authors: Chi Kit Wong, Yan Liu, Haowen Yan

Abstract: We present a brain-to-image system that decodes visual stimuli from EEG signals recorded during natural image viewing. Our system addresses two tasks: (1) EEG-to-image retrieval, which ranks the correct stimulus image among 200 candidates given an EEG segment, and (2) EEG-to-image reconstruction, which generates an image consistent with the perceived stimulus. For retrieval, we implement a multi-l… ▽ More We present a brain-to-image system that decodes visual stimuli from EEG signals recorded during natural image viewing. Our system addresses two tasks: (1) EEG-to-image retrieval, which ranks the correct stimulus image among 200 candidates given an EEG segment, and (2) EEG-to-image reconstruction, which generates an image consistent with the perceived stimulus. For retrieval, we implement a multi-level blurring approach improved with biologically inspired EVNet features and trained with the InfoNCE loss. Evaluated over 10 random seeds for a single subject, the retrieval model achieves a mean final-epoch Top-1 accuracy of 86.30% and Top-5 accuracy of 98.55%. For reconstruction, we implement CognitionCapturerPro, which aligns EEG representations to multi-modal CLIP embeddings, including image, text, depth, and edge embeddings, and synthesizes images with SDXL-Turbo conditioned via IP-Adapter. Averaged over 10 seeds, the reconstruction model achieves a CLIP score of 0.903 using ViT-H-14, a CLIP score of 0.870 using ViT-L/14, and an SSIM of 0.409. These results demonstrate the feasibility of decoding rich visual representations from EEG signals using modern multi-modal alignment and generative modeling techniques. △ Less

Submitted 18 May, 2026; originally announced May 2026.

Comments: 16 pages, 5 figures. Code available at: https://github.com/Chikit-WONG/DL_Project/

Fluid RIS (FRIS)-Assisted Index Modulation for 6G Wireless Communications

Authors: Xusheng Zhu, Kai-Kit Wong, Sai Xu, Hao Xu, Wen Chen, Hyundong Shin

Abstract: Fluid reconfigurable intelligent surfaces (FRIS) extend conventional reconfigurable intelligent surfaces (RIS) by adding spatial reconfigurability through switchable apertures, pattern-reconfigurable units, fluidic conductive materials, or movable surface elements. This article studies how FRIS can support index modulation (IM), where information bits select a surface configuration and the receive… ▽ More Fluid reconfigurable intelligent surfaces (FRIS) extend conventional reconfigurable intelligent surfaces (RIS) by adding spatial reconfigurability through switchable apertures, pattern-reconfigurable units, fluidic conductive materials, or movable surface elements. This article studies how FRIS can support index modulation (IM), where information bits select a surface configuration and the receiver detects the index from the induced receiver-side response. A key challenge is that many feasible FRIS layouts do not necessarily lead to many reliable spatial indices. After propagation, mutual coupling, hardware distortion, and receiver observation, different layouts may produce similar receiver-side responses and cause index-detection errors. To address this issue, we present a response-aware design view, in which FRIS spatial codebooks are selected according to response-domain separability rather than layout diversity alone. We also discuss actuation granularity as a practical design knob that balances spatial diversity, pilot overhead, coupling robustness, and hardware feasibility. The resulting workflow helps select compact, trainable, and controllable spatial-index codebooks from dense FRIS layouts, providing design guidance for future programmable wireless environments. △ Less

Submitted 21 May, 2026; originally announced May 2026.

Finite-Aperture Planar Fluid Antenna Array

Authors: Zhentian Zhang, Jingyuan Xu, Kai-Kit Wong, Hao Jiang, Zaichen Zhang, Hyundong Shin

Abstract: Fluid antenna systems (FASs) are emerging as a reconfigurable-aperture technology that expands physical-layer design beyond fixed, rigid antenna geometries. While the \emph{fading diversity} of FASs -- which exploits spatial channel fluctuations for signal enhancement and interference avoidance -- has been widely studied, the \emph{geometry diversity} created by reconfigurable port placement remai… ▽ More Fluid antenna systems (FASs) are emerging as a reconfigurable-aperture technology that expands physical-layer design beyond fixed, rigid antenna geometries. While the \emph{fading diversity} of FASs -- which exploits spatial channel fluctuations for signal enhancement and interference avoidance -- has been widely studied, the \emph{geometry diversity} created by reconfigurable port placement remains far less understood, particularly for planar architectures under finite-aperture constraints. This paper develops a systematic analytical framework for finite-aperture planar fluid antenna arrays (FAAs). First, we derive a closed-form characterization of the minimum inter-port distance under uniform random placement over a rectangular aperture and show that it follows a Rayleigh law. Its mean scales as $\mathcal{O}(M^{-1})$, in sharp contrast to the $\mathcal{O}(M^{-2})$ behavior in the linear case in which $M$ represents the number of candidate ports, revealing a fundamentally more favorable packing geometry in two dimensions. Secondly, we establish a universal Cramér-Rao bound (CRB) for joint elevation-azimuth estimation, governed by a $2\times 2$ \emph{geometric inertia matrix} whose determinant and eigenstructure fully capture the role of port placement in estimation precision. We further prove that both the trace and determinant of this matrix are invariant to the azimuth look direction. Third, we uncover an intrinsic \emph{precision--ambiguity trade-off}: maximizing the geometric determinant to minimize the CRB drives ports toward the aperture boundary, but simultaneously increases sidelobe-induced spatial ambiguity. △ Less

Submitted 21 May, 2026; originally announced May 2026.

Industrial Dual-Arm Box Handling via Online Inertial Estimation and Convex Wrench Optimization

Authors: Kenzhi Iskandar Wong, Lin Yang, Qian Ying Lee, Domenico Campolo

Abstract: Industrial robotic object handling often involves boxes and packages whose mass and center of mass are not known in advance. These uncertainties affect the force--moment balance required for stable lifting, and improper regulation of contact wrenches can lead to slip, object drop, orientation deviation, or excessive squeezing. This paper presents a friction-aware dual-arm box-handling framework fo… ▽ More Industrial robotic object handling often involves boxes and packages whose mass and center of mass are not known in advance. These uncertainties affect the force--moment balance required for stable lifting, and improper regulation of contact wrenches can lead to slip, object drop, orientation deviation, or excessive squeezing. This paper presents a friction-aware dual-arm box-handling framework for objects with unknown inertial properties. The proposed approach estimates the object mass and center of mass online from measured contact wrenches, and computes friction-feasible contact forces and torsional moments through a second-order cone program (SOCP) under ellipsoidal friction-limit-surface constraints. An offline trajectory refinement stage is also included to reduce undesired object--environment contact when geometric constraints are present. By enforcing friction feasibility as a hard constraint and minimizing contact effort within the feasible region, the framework achieves stable lifting without treating slip avoidance and excessive squeezing as separately tuned objectives. Experiments on a real dual-arm robotic system under different center-of-mass configurations demonstrate that the method lifts objects with unknown inertial properties while maintaining stable frictional contact. △ Less

Submitted 21 May, 2026; originally announced May 2026.

Comments: 14 pages, submitted to Robotics and Computer-Integrated Manufacturing (RCIM) Journal

PrivacyAkinator: Articulating Key Privacy Design Decisions by Answering LLM-Generated Multiple-choice Questions

Authors: Qiyu Li, Yuen Sum Wong, Yuen Kei Wong, Longxuan Yu, Haojian Jin

Abstract: NIST's Privacy Risk Assessment Methodology (PRAM) provides a structured framework for privacy experts to assess privacy risks. However, its complexity and reliance on expert knowledge make it difficult for novice developers to use effectively. This paper explores methods to lower these barriers. We first performed an observational study with 12 participants using PRAM in real-world scenarios, and… ▽ More NIST's Privacy Risk Assessment Methodology (PRAM) provides a structured framework for privacy experts to assess privacy risks. However, its complexity and reliance on expert knowledge make it difficult for novice developers to use effectively. This paper explores methods to lower these barriers. We first performed an observational study with 12 participants using PRAM in real-world scenarios, and found that novice developers struggled most with articulating privacy-related design decisions. We then developed PrivacyAkinator, an interactive tool that helps developers articulate key privacy decisions by answering LLM-generated multiple-choice questions. PrivacyAkinator introduces three innovations: a universal privacy representation that abstracts privacy-related design decisions into data flows and stakeholder interactions; a domain-aware design space mined from 10K privacy-related news articles; and a dynamic question-generation workflow to prioritize relevant questions. Our user study with 24 participants suggests that developers using PrivacyAkinator identified 47% more key decisions in 73% less time compared to PRAM. △ Less

Submitted 7 April, 2026; originally announced May 2026.

Comments: Accepted to ACM CHI 2026

Sonalyzer-Moz: A Framework for Analyzing the Structure of Mozart's Sonata Form

Authors: Jing Zhao, KokSheik Wong, Vishnu Monn Baskaran, Kiki Adhinugraha, David Taniar

Abstract: The sonata form is a musically rich and hierarchically structured form that poses significant challenges for automatic analysis. While music structure analysis has seen strides of progress in recent years, sonata form analysis remains in its early stages. This is largely due to the time-consuming and high barrier of the music background requirement for annotating classical music structures. To adv… ▽ More The sonata form is a musically rich and hierarchically structured form that poses significant challenges for automatic analysis. While music structure analysis has seen strides of progress in recent years, sonata form analysis remains in its early stages. This is largely due to the time-consuming and high barrier of the music background requirement for annotating classical music structures. To advance research in this area, we curated SoSA-Moz, the first large-scale dataset featuring comprehensive hierarchical structure annotations. This work establishes a foundation for systematic sonata form analysis. Leveraging this newly contributed resource, we further propose Sonalyzer-Moz, a baseline model specifically designed for investigating complex sonata structures. This framework integrates feature aggregation with sequential modeling, enabling it to capture both local feature and upper-level structural dependencies. Experiment results show that Sonalyzer-Moz is capable of identifying the components' boundaries of the upper-level structure that are critical to understanding sonata form. Therefore, this method demonstrates, for the first time, the effectiveness of automatic upper-level analysis of sonata form, and provides a robust baseline for future research in the automatic understanding of sonata form while advancing the study of classical music structure analysis. △ Less

Submitted 14 August, 2026; v1 submitted 18 May, 2026; originally announced May 2026.

Comments: 6 pages, 2 figures, Accepted at APSIPA ASC 2026. Camera-ready version

Beyond Localization: A Comprehensive Diagnosis of Perspective-Conditioned Spatial Reasoning in MLLMs from Omnidirectional Images

Authors: Yuangong Chen, Wai Keung Wong, Jiaxing Li, Ioannis Patras, Xu Zheng

Abstract: Multimodal Large Language Models (MLLMs) show strong visual perception, yet remain limited in reasoning about space under changing viewpoints. We study this challenge as Perspective-Conditioned Spatial Reasoning (PCSR) in 360-degree omnidirectional images, where broad scene coverage reduces ambiguity from partial observations without eliminating the need for viewpoint-dependent inference. To asses… ▽ More Multimodal Large Language Models (MLLMs) show strong visual perception, yet remain limited in reasoning about space under changing viewpoints. We study this challenge as Perspective-Conditioned Spatial Reasoning (PCSR) in 360-degree omnidirectional images, where broad scene coverage reduces ambiguity from partial observations without eliminating the need for viewpoint-dependent inference. To assess this capability, we introduce PCSR-Bench, a diagnostic benchmark of 84,373 question-answer pairs from 2,600 omnidirectional images across 26 indoor environments. PCSR-Bench contains eight tasks spanning foundational perception (e.g., object counting, relative distance, and relative direction) and advanced PCSR, including compositional chains, egocentric rotation, perspective re-anchoring, ego-distortion, and limited-FOV visibility. We evaluate 14 representative MLLMs and observe a substantial perception-reasoning gap: accuracy reaches 57.59% on foundational relative direction, but drops to 13.49% on egocentric rotation, 7.13% on egocentric distortion, and 0.64% on open-ended compositional reasoning. To probe the plasticity of this gap, we conduct an RL-based diagnostic study on a 7B-scale model. Reward shaping improves a matched 7B baseline from 31.10% to 60.06% under a controlled setting, suggesting that PCSR is partial plasticity rather than being fully immutable. Still, the gains are task-selective, sensitive to reward design including both weight allocation and reward formulation, and partially dependent on the evaluation protocol. These results position PCSR as a key bottleneck in current MLLMs and highlight limited but meaningful room for recovery under targeted optimization. △ Less

Submitted 18 May, 2026; v1 submitted 12 May, 2026; originally announced May 2026.

Comments: 10pages, 4 figures

ACM Class: I.2.10; I.2.7; I.4.8

Traffic Scenario Orchestration from Language via Constraint Satisfaction

Authors: Frieda Rong, Chris Zhang, Kelvin Wong, Raquel Urtasun

Abstract: Autonomous vehicles (AVs) require extensive testing in simulation, but test case generation for driving scenarios is laborious. The desired scenarios are often out-of-distribution and have precise requirements on interactions with the AV policy under test. Manually programming scenarios allows for precise controllability but is difficult to scale. On the other hand, statistical models can leverage… ▽ More Autonomous vehicles (AVs) require extensive testing in simulation, but test case generation for driving scenarios is laborious. The desired scenarios are often out-of-distribution and have precise requirements on interactions with the AV policy under test. Manually programming scenarios allows for precise controllability but is difficult to scale. On the other hand, statistical models can leverage compute and data, but struggle with precise controllability when out-of-distribution. We cast scenario orchestration as a constraint-solving problem and present a language-in, simulation-out scenario orchestrator for closed-loop testing AVs. Our approach leverages foundation model reasoning to translate general, natural language descriptions into a set of constraints as a scenario representation. This then allows us to leverage off the shelf solvers to solve for actor behaviors which meet precise testing intentions in closed-loop. Under a benchmark of carefully crafted and diverse scenario descriptions, our approach greatly outperforms our baselines in orchestration success rate. We further show that our closed-loop approach is especially important for scenarios which require ego-reactive specifications. △ Less

Submitted 7 May, 2026; originally announced May 2026.

Comments: 19 pages, 10 figures; full version of paper accepted for poster presentation at ICRA 2026

Fluid Antenna Systems Enabling 6G HRLLC With Port Switching Delay

Authors: Xusheng Zhu, Kai-Kit Wong, Hao Xu, Chenguang Rao, Hyundong Shin

Abstract: Fluid antenna systems (FAS) exploit antenna position reconfigurability to unlock massive spatial diversity within compact form factors, making them a promising enabler for 6G user terminals (UTs). However, practical port switching incurs latency and signaling overhead, which can be particularly detrimental to hyper-reliable low-latency communications (HRLLC) under finite blocklength operation. Thi… ▽ More Fluid antenna systems (FAS) exploit antenna position reconfigurability to unlock massive spatial diversity within compact form factors, making them a promising enabler for 6G user terminals (UTs). However, practical port switching incurs latency and signaling overhead, which can be particularly detrimental to hyper-reliable low-latency communications (HRLLC) under finite blocklength operation. This paper investigates FASenabled HRLLC by explicitly capturing the coupled effects of spatial correlation, port switching delay, and finite blocklength coding. We derive exact closed-form expressions for the average block error rate (BLER) and average achievable rate over spatially correlated fading channels. The resulting analysis reveals a fundamental design trade-off: increasing the number of ports improves diversity but linearly reduces the effective blocklength, thereby intensifying finite-blocklength penalties. A key theoretical contribution is a rigorous proof that reliability, achievable rate, and energy efficiency are strictly unimodal in the port dimension, ensuring a unique optimal port configuration. Furthermore, we characterize an explicit switching-delay threshold that separates regimes where FAS yields net gains over fixed-position antenna (FPA) systems. Numerical results validate the analysis and show that substantial HRLLC performance gains are achievable when the switching latency remains below the derived bound. △ Less

Submitted 9 June, 2026; v1 submitted 7 May, 2026; originally announced May 2026.

Phased Ultra Massive Array (PUMA)

Authors: Hanjiang Hong, Kai-Kit Wong, Xusheng Zhu, Chenguang Rao, Dazhi He, Hyundong Shin

Abstract: This paper proposes a novel multiple-access framework, termed the phased ultra massive antenna array (PUMA), which exploits the distinctive spatial flexibility of fluid antenna systems (FAS) at the user equipment (UE). Building upon fluid antenna multiple access (FAMA) and compact ultra-massive antenna array (CUMA), PUMA incorporates a phased array for signal aggregation. This architecture enables… ▽ More This paper proposes a novel multiple-access framework, termed the phased ultra massive antenna array (PUMA), which exploits the distinctive spatial flexibility of fluid antenna systems (FAS) at the user equipment (UE). Building upon fluid antenna multiple access (FAMA) and compact ultra-massive antenna array (CUMA), PUMA incorporates a phased array for signal aggregation. This architecture enables the UE to inherently mitigate co-user interference within the spatial domain without necessitating channel state information (CSI) for precoding at the base station (BS) or complex interference cancellation at each UE. A primary advantage of PUMA lies in its hardware efficiency: by implementing phase shifting and signal combining in the analog domain, it achieves high antenna gain while requiring only a minimal number of radio-frequency (RF) chains, potentially a single RF chain. Comprehensive theoretical analysis of the achievable data rate is provided, complemented by extensive simulations that validate the framework. The results demonstrate that PUMA markedly outperforms FAMA and CUMA architectures, particularly for UEs with a single RF chain, offering a robust and scalable solution for interference-insensitive massive connectivity in sixth-generation (6G) systems. △ Less

Submitted 6 May, 2026; originally announced May 2026.

Conditional Flow-VAE for Safety-Critical Traffic Scenario Generation

Authors: Zimu Gong, Brian Zhaoning Zhang, Chris Zhang, Kelvin Wong, Raquel Urtasun

Abstract: Safety-critical scenarios are essential for the development of autonomous vehicles (AVs) but are rare in real-world driving data. While simulation offers a way to generate such scenarios, manually designed test cases lack scalability, and adversarial optimization often produces unrealistic behaviors. In this work, we introduce a conditional latent flow matching approach for scalable and realistic… ▽ More Safety-critical scenarios are essential for the development of autonomous vehicles (AVs) but are rare in real-world driving data. While simulation offers a way to generate such scenarios, manually designed test cases lack scalability, and adversarial optimization often produces unrealistic behaviors. In this work, we introduce a conditional latent flow matching approach for scalable and realistic safety-critical scenario generation. Our method uses distribution matching to transform nominal scenes into safety-critical rollouts. Furthermore, we demonstrate that incorporating both simulation and real-world data enables our framework to efficiently generate diverse, data-driven scenarios. Experimental results highlight that our approach is able to more consistently and realistically generate novel safety-critical scenarios, making it a valuable tool for training and benchmarking AV systems. △ Less

Submitted 5 May, 2026; originally announced May 2026.

Comments: ICRA 2026

MAGRPO: Accelerated MARL Training for Fluid Antenna-Assisted Wireless Network Optimization

Authors: Wanzhe Wang, Tong Zhang, Hao Xu, Shuai Wang, Rui Wang, Kai-Kit Wong

Abstract: Fluid antenna system (FAS) becomes a promising paradigm for next-generation wireless networks, which enables position-flexible antenna elements that can dynamically adjust to more favorable channel conditions. However, the optimization of fluid antenna (FA) positions, beamforming, and power allocation in FA-assisted wireless networks is challenging, due to the non-convexity and the lack of base st… ▽ More Fluid antenna system (FAS) becomes a promising paradigm for next-generation wireless networks, which enables position-flexible antenna elements that can dynamically adjust to more favorable channel conditions. However, the optimization of fluid antenna (FA) positions, beamforming, and power allocation in FA-assisted wireless networks is challenging, due to the non-convexity and the lack of base station (BS) coordination. In this paper, we first formulate this challenging optimization problem as a decentralized partially observable Markov decision process, and then propose a multi-agent group relative policy optimization (MAGRPO) algorithm under the centralized training decentralized execution (CTDE) paradigm. Compared with multi-agent proximal policy optimization (MAPPO), MAGRPO replaces the critic network with group relative advantage estimation. This design reduces computational complexity by nearly half under parameter sharing. Furthermore, we derive a variance upper bound of the cumulative reward, which scales with network parameters, e.g., the number of BSs, users, and FAs. Simulation results show that compared with wireless networks with fixed antenna positions, FA-assisted wireless networks achieve multiple-fold sum-rate enhancement. Moreover, the proposed MAGRPO attains sum-rates comparable to those of MAPPO in testing, while reducing training time by $30\% \sim 40\%$. △ Less

Submitted 1 May, 2026; v1 submitted 19 April, 2026; originally announced April 2026.

Comments: 13 pages,9 figures

Jointly Correlated Dual-Side Fluid Antenna System

Authors: Zhentian Zhang, Yuanhui Wu, Kai-Kit Wong, Hao Jiang, An Li

Abstract: Fluid antenna systems (FASs) have introduced a new paradigm for wireless system design by revealing how mutual correlation can be exploited to harvest inherent spatial diversity. While existing studies have mainly focused on one-sided FAS configurations, i.e., with FAS deployed at either the transmitter or the receiver, this work investigates the ergodic capacity of a jointly correlated dual-side… ▽ More Fluid antenna systems (FASs) have introduced a new paradigm for wireless system design by revealing how mutual correlation can be exploited to harvest inherent spatial diversity. While existing studies have mainly focused on one-sided FAS configurations, i.e., with FAS deployed at either the transmitter or the receiver, this work investigates the ergodic capacity of a jointly correlated dual-side FAS under statistical eigenmode transmission. Specifically, a jointly correlated dual-side channel model is developed, and the corresponding ergodic capacity together with a tight closed-form upper bound is derived. In addition, the optimal power allocation is studied, and a practical iterative algorithm is proposed for its implementation. △ Less

Submitted 18 April, 2026; originally announced April 2026.

Beyond Covariance: Generative Spatial Correlation Modeling and Channel Interpolation for Fluid Antenna Systems

Authors: Zhentian Zhang, Hao Jiang, Kai-Kit Wong, Hyundong Shin, Ross Murch

Abstract: Fluid antenna systems (FAS) enable unprecedented spatial diversity within a compact form factor by flexibly switching among high-density antenna ports. To activate this capability, channel state information (CSI) over the ports is required, which implies high estimation overhead because the number of ports is usually very large. Conventional estimation schemes tend to first estimate the CSI for a… ▽ More Fluid antenna systems (FAS) enable unprecedented spatial diversity within a compact form factor by flexibly switching among high-density antenna ports. To activate this capability, channel state information (CSI) over the ports is required, which implies high estimation overhead because the number of ports is usually very large. Conventional estimation schemes tend to first estimate the CSI for a small number of ports and then infer the CSI for the remaining antenna ports by interpolation exploiting correlation characteristics. However, existing correlation-based techniques lack generalization ability, and the fundamental limits of interpolating the CSI from sparse observations remain poorly understood. This paper adopts a generative modeling framework for characterizing the channel correlation among the FAS ports that departs fundamentally from covariance-descriptive models. Specifically, we represent the spatially sampled channel as a $p$th-order autoregressive (AR) Gauss-Markov process, which provides a principled and tunable tradeoff between model complexity and approximation accuracy via the AR order. In so doing, we can characterize the limits of channel interpolation by deriving the globally optimal minimum mean-square error (MMSE) estimator and establishing a tight lower bound on the minimum number of observations required to meet a prescribed reconstruction error. To reduce the complexity of MMSE estimation, we then exploit the state-space structure due to the ${\rm AR}(p)$ model and develop a Kalman filtering/smoothing-based interpolation algorithm. The resulting method attains the optimal MMSE performance with strictly linear complexity $\mathcal{O}(N)$ with $N$ denoting the number of ports, resulting in a scalable, efficient, and theoretically grounded framework for practical FAS channel reconstruction. △ Less

Submitted 17 April, 2026; originally announced April 2026.