Searcharxiv⌕ Search

arXiv subjects

Ziang Liu

Publications and source records attributed to Ziang Liu.

At least 19 recordsLinked to original sources

AcousticBEV: Acoustic BEV Object Localization for Intelligent Vehicles

Modern autonomous driving and driver assistance systems rely predominantly on cameras, radars and LiDARs for environmental perception. However, these modalities face different environmental limitations, including poor illumination for cameras and adverse weather for LiDAR. Acoustic sensing, particularly using microphone arrays, could provide complementary perception capabilities for autonomous driving, but has remained largely underexplored. In this work, we design a horizontally arranged microphone array that functions as the "ears" of a vehicle, enabling localization of nearby moving vehicles in a bird's-eye view (BEV) representation that captures both their relative direction and distance. Using this array, we collect a real-world driving dataset, AcousticScenes, which contains 2,117 static and 2,863 dynamic samples with synchronized spatial acoustic signals, LiDAR point clouds and 3D bounding box annotations. Building on this dataset, we propose AcousticBEV, a learning-based method for acoustic BEV object localization, which estimates the center locations of moving traffic participants in BEV from the microphone-array signals, conditioned on the ego vehicle's own speed. Experimental results demonstrate that AcousticBEV outperforms both classical signal-processing baselines and other learning-based methods, achieving F1 scores of $47.7 \pm 1.2$% on "Easy" scenarios and $45.1 \pm 1.4$% on "Difficult" cases. Our findings demonstrate that acoustic signals from a low-cost, vehicle-mounted microphone array can localize nearby moving vehicles in both azimuth and distance.

cs.CV↗

Can STATCOM and UPS-Based Voltage Support Mitigate Fault-Induced Reductions in AI Data Center Grid Power Draw?

Artificial intelligence data centers (AIDCs) are rapidly increasing in number and capacity. These large electronic loads are interfaced through power electronic devices and can sharply reduce grid power intake during fault-induced voltage sags. Disturbances exceeding equipment ride-through capabilities may also trigger protective disconnection, posing risks to power system security. Static synchronous compensators (STATCOMs) and appropriately controlled AIDC uninterruptible power supply (UPS) rectifiers can provide reactive power support during voltage sags. This paper compares the individual and combined effectiveness of STATCOM and UPS-based voltage support in mitigating fault-induced reductions in grid power intake through electromagnetic transient simulations of a 1 GW AIDC connected to an IEEE nine-bus system. Simulation results show that STATCOM support better preserves short-term intake in milder tested faults, whereas UPS coordination maintains rectifier operation under deeper sags by reducing active intake to reserve reactive current capacity. Combined support improves auxiliary load retention and recovery in selected cases. However, severe tested sags still cause substantial grid-demand reductions even with combined support.

eess.SY↗

AgentSSL: Can MLE Agents Leverage Unlabeled Data?

Applying machine learning systems to new domains introduces bottlenecks in both labeled data acquisition and manual pipeline engineering. While recent MLE agents show promise in automating pipeline design, they have been studied almost exclusively in fully supervised settings, leaving them untested in realistic regimes where labeled data is scarce but unlabeled data is abundant. In this work, we introduce Agentic Semi-Supervised Learning (AgentSSL), framing SSL as a program search problem and studying how agents navigate this space under limited supervision. We find that even under weak feedback and matched compute, agents can synthesize code that combines and adapts established SSL techniques, exceeding the strongest human-engineered baselines in 8 of 12 settings by up to +21.8%. However, our results identify program selection as a critical bottleneck in AgentSSL, where performance drops by as much as 47% when selecting based on the agent's feedback metric. To systematically study this bottleneck, we introduce ProgramSelectBench, a large-scale benchmark with a combined total of 16,977 agent-generated and human-engineered SSL programs, and evaluate both unsupervised and active program selection methods. We further evaluate AgentSSL on real-world ecology detection tasks, highlighting the opportunities and challenges of transferring agent-discovered strategies beyond controlled benchmarks.

cs.AI↗

Data-Driven Robust Dynamic Wireless Charging Lane Planning with Tunable Profit--Risk Trade-off and Modified User Equilibrium

Dynamic wireless charging (DWC) has attracted increasing attention as a complement to conventional plug-in charging for electric vehicles (EVs). However, existing planning methods for dynamic wireless charging lanes (DWCLs) overlook realistic EV routing and charging behavior and lack explicit out-of-sample risk control under traffic-demand and renewable-energy uncertainties. To address these limitations, this paper proposes a data-driven two-stage robust planning framework that coordinates DWCL deployment with the operation of coupled traffic--power networks (CTPNs). A data-driven uncertainty set is constructed from joint historical scenarios to preserve spatio-temporal correlations without imposing parametric distributional assumptions. A modified user-equilibrium (MUE) model is proposed to capture realistic EV routing and charging behavior by incorporating state-of-charge (SoC) dynamics and endogenous charging decisions. The resulting two-stage robust model with mixed-integer recourse is reformulated as an equivalent single-stage scenario-approach problem and solved using a customized nested column-and-constraint generation algorithm. An \textit{a posteriori} feasibility guarantee is established to quantify out-of-sample risk. Based on this guarantee, the scenario set is iteratively enriched to meet a prescribed risk tolerance. Case studies on a 69-node power distribution network and a 24-node traffic network demonstrate that the proposed MUE model prevents overinvestment in DWCLs, while the proposed risk-quantification procedure enables a tunable profit--risk trade-off.

eess.SY↗

Contrastive Discovery: Open-Ended Scientific Discovery over Competing Explanations

LLM agents are increasingly used for open-ended, data-driven scientific discovery, yet most of them evaluate hypotheses pointwise, testing an empirical claim $h$ only against its complement $\neg h$. When evidence rejects $h$, the agent learns that one explanation is unlikely but not which alternative the evidence supports. To recover that alternative, we introduce Contrastive Discovery, which replaces the single hypothesis with a contrast set, a scientific question paired with mutually exclusive and collectively exhaustive competing explanations. Each experiment is scored by how its evidence redistributes belief across the set, and the question is resolved when belief concentrates on one explanation. Because contrast sets change only the unit an agent proposes and scores, an existing agent can use them to re-evaluate evidence it has already collected or to guide its search. Across six domains, $89\%$ of the pointwise candidates with the largest belief shifts are rejections. On the same evidence, contrast sets resolve $73\%$ of these candidates to a specific alternative. When contrast sets guide search, the agent resolves $67\%$ more questions than pointwise surprisal under the same experiment budget. In a blinded study, $38$ human experts chose the same explanation on $78\%$ of resolved questions and on $82\%$ of questions where the pointwise agent rejected its hypothesis and stopped. Together, these results support explicit competing explanations as a way for discovery agents to resolve more scientific questions from the same evidence and validation budget.

cs.AI↗

ScienceIDE: Turning World's Scientific Codebase into Agent Learnable Environments

Scientific code repositories encode decades of human knowledge in executable models, methods, and tools. Yet fragmented toolchains, implicit domain conventions, and specialized correctness criteria make this knowledge difficult to convert into reliable learning experience-a challenge we call the scientific experience bottleneck. We introduce ScienceIDE, infrastructure for turning the world's scientific code into programmable environments for scientific agents. Guided by expert-defined scientific cases and acceptance criteria, agents transform repositories into executable environments that support task generation, execution, and scientific verification. These environments provide a shared foundation for supervised fine-tuning, reinforcement learning, and evaluation. Using verified interaction trajectories, we train PhAI-IDE-72B/9B/4B. The model family shows gains in held-out scientific-code repair and across selected general-purpose benchmarks in code, reasoning, and knowledge, providing evidence of positive transfer from scientific experience to broader capabilities. ScienceIDE lays the foundation for an integrated workspace for agent learning and scientific practice, making humanity's scientific software a shared substrate for developing scientific intelligence. Code: https://github.com/aitofound/ScienceIDE

cs.CL↗

HistoLoop: Iterative Contextual Inference with a Looped Transformer for Histology-Based Spatial Gene Expression Prediction

Predicting spatial gene expression from H&E images offers a scalable complement to experimental spatial transcriptomics. However, despite progress in spatial context modeling, how to construct subsequent context from spot states updated by earlier contextual interactions remains underexplored. We introduce HistoLoop, a looped Transformer that formulates prediction as iterative contextual inference. At each refinement step, HistoLoop aggregates the current spot states into context tokens at the corresponding resolution, processes them with a shared spatial Transformer, and retrieves spot-specific contextual information through cross-attention. A memory-mediated update integrates this information into persistent spot states, which are then used to construct context at the next, finer resolution. Under the HEST-Bench HVG50 evaluation protocol, HistoLoop achieves the highest macro-averaged gene-wise Pearson correlation among the compared methods across six cancer-specific tasks spanning five organs in HEST-1K. Ablation studies support the importance of recurrent state refinement and reconstructing context from updated spot states at progressively finer resolutions. Expression-derived clustering further suggests that HistoLoop better captures spatially organized multigene patterns.

cs.CV↗

Data Center Load Accommodation under QoS and Checkpoint-Induced Frequency Security Constraints

The rapid growth of artificial intelligence workloads is accelerating the expansion and grid interconnection of data centers (DCs). Existing data center load accommodation (DCLA) assessments primarily focus on grid and resource adequacy and often imply that greater workload flexibility always increases accommodable capacity. However, workload flexibility has a dual effect. Workload deferral and temporal shifting can alleviate network congestion, whereas synchronized checkpointing of training workloads can cause large and rapid load reductions, threatening frequency stability. Therefore, DCLA cannot be determined solely by grid adequacy. This paper develops an optimal DCLA method that jointly accounts for workload quality of service (QoS), checkpoint-induced frequency security (FS), and grid adequacy. First, a QoS-aware model characterizes workload-dependent flexibility while ensuring service feasibility. Second, checkpoint-induced FS constraints are developed for periodic and event-triggered power variations, with key frequency-security limits reformulated in algebraic form. Third, a two-stage lexicographic model is developed to maximize DCLA and minimizes the required energy storage (ES) deployment, and then optimizes system operation under the given DC and ES siting and sizing decisions. Case studies show that the proposed method avoids a 40.3% overestimation of DCLA compared with the grid-adequacy benchmark. QoS is the primary limiting factor, while checkpoint-induced frequency-security constraints become more restrictive as training workloads increase. These findings highlight that reliable DCLA should account for both its grid-support value and the associated QoS and FS risks.

eess.SY↗

VETNET: Assembling Benchmark Datasets for AI-Driven Veterinary Innovation

VETNET represents an ecosystem of live benchmark datasets and evaluation frameworks designed to accelerate AI-driven innovation in animal science and veterinary medicine through standardized, high-quality datasets and multi-institutional collaboration across academic and non-academic sectors. This framework addresses the gap between AI advances in human medicine and the fragmented state of veterinary AI development.

cs.AI↗

Privacy-Preserving Coordinated Operation of Power Grids and AI Data Centers: A Checkpoint-Aware Three-Phase Scheme

The rapid growth of large language model training and serving is driving AI data centers (AIDCs) toward gigawatt scale. Unlike conventional commercial loads, AIDCs possess significant operational flexibility through dynamic voltage and frequency scaling (DVFS) of training and inference workloads, while periodic model checkpointing can induce abrupt power drops and rebounds that erode operating reserves and increase transmission congestion risks. Coordinating AIDC operation with grid scheduling under these unique operational characteristics is challenging because grid and AIDC operators are generally unwilling to share proprietary data and decision-making authority. This paper proposes a hierarchical privacy-preserving coordinated operation scheme between the power grid and AIDCs to address this gap. The proposed scheme contains three phases. In Phase I, the grid operator computes a certified inner approximation of the AIDCs security region for subsequent coordination. In Phase II, the AIDC operator coordinates training and inference AIDCs to optimize workload allocation within the certified security region and generate power schedules and checkpoint alerts. In Phase III, the grid operator solves a checkpoint-aware two-stage robust optimal power flow (OPF) considering renewable generation and checkpoint uncertainties. By exchanging only compact interface information, the framework preserves the privacy of both grid and AIDCs, avoids frequent iterative communication, and enables secure coordination with guaranteed feasibility. Numerical studies on a modified IEEE 14-bus system and a modified NYISO system demonstrate the effectiveness, robustness, and security of the proposed framework.

eess.SY↗

WildFin: An In-the-Wild Dataset for Fish Behavioral Recognition

Recent advances in field technology have led to a massive influx of in-the-wild video data for ecological science. The primary bottleneck in leveraging this data is the high cost of expert annotation. While computer vision offers a potential solution, current models frequently fail when deployed in complex marine environments. To characterize these failures, we introduce WildFin, a novel benchmark for fish behavior recognition collected and annotated by ecologists. WildFin spans two critical real-world paradigms: stationary cameras monitoring groups of fish and dynamic divers following individual subjects. The dataset represents a massive curation effort, involving 1,350 hours of fieldwork and 600 hours of expert annotation to produce 9 hours of behavioral data with over 2 million frame-by-frame labels. We benchmark modern vision foundation models and quantify tradeoffs between static and spatiotemporal architectures, revealing the substantial gap that remains between current model capabilities and the demands of real-world underwater behavioral analysis. Project website: https://team-wildfin.github.io/.

cs.CV↗

HistoGPA: A Context-Conditioned Gene-Prior Attention Framework for Histology-Based Spatial Gene Expression Prediction

Predicting spatial gene expression from routine hematoxylin and eosin (H&E) images provides a practical complement to experimental spatial transcriptomics. Existing approaches focus on local or multi-scale visual features and often treat pretrained gene representations as fixed priors, although the interpretation of local morphology and the relevance of gene priors depend on tissue context. We propose HistoGPA, a context-conditioned gene-prior attention framework that uses a shared slide-level representation in two parallel pathways: one modulates local morphological features, whereas the other conditions pretrained gene embeddings and retrieves gene-prior information through cross-attention. This design enables each spatial location to retrieve context-adapted gene-prior information using its local morphology, position, and slide context. Across ten cancer types in HEST-1k, HistoGPA achieves the highest macro-averaged gene-wise Pearson correlation coefficient among the compared methods under the same evaluation protocol for both the top-50 and top-1,500 highly variable gene sets. Additional analyses show that HistoGPA better recovers the spatial expression patterns of cancer-associated genes and yields greater agreement between clusters derived independently from predicted and ground-truth expression profiles. Together, these findings motivate a context-dependent view of histology-to-expression prediction, in which local morphological representations and gene priors are jointly adapted to the broader tissue context.

cs.CV↗

From Chatbot to Digital Colleague: The Paradigm Shift Toward Persistent Autonomous AI

Large Language Models (LLMs) are undergoing a fundamental transformation from conversational generators into integrated AI systems capable of reasoning, action, memory, and self-improvement. We conceptualize this transition as a shift from Chatbot to Digital Colleague: from conversational answers to persistent work. We organize this transition along two tightly coupled dimensions. First, at the cognitive core level, LLMs are advancing from Chatbot-era "fast thinking" systems driven by next-token prediction toward Thinking LLMs that leverage inference-time computation, Chain-of-Thought reasoning, reflection, process supervision, and reinforcement learning to support more deliberate and reliable cognition. Second, at the tool-augmented task execution level, LLMs are progressing from tool-calling Agents that invoke external resources in an ad hoc manner toward OpenClaw-style workstation systems (OpenClaw) equipped with persistent Workspaces, skills, verification loops, and governance. The "Workspace + Skill" paradigm makes episodic tool use colleague-like via state persistence, reusable procedures, task closure, and experience reuse. We examine data construction shifts from instruction-response pairs to State-Action-Observation trajectories and evaluation from static benchmarks to sandboxed, auditable, self-evolving AI ecosystems.

cs.AI↗

Thinking in Video: Can Video Generators Really Reason About the Real World?

Recent advances in world models and video generation have given rise to an emerging reasoning paradigm that leverages video generative models to simulate, predict, and reason about real-world dynamics. We redefine this paradigm as Thinking in Video, where video is not merely an output artifact but a medium for constructing, extending, and verifying causal thought. However, this promise remains unverified: convincing rollouts may reflect memorized appearances rather than causal understanding, while existing metrics separate perceptual fidelity from semantic logic. To evaluate whether video generators support such reasoning, we introduce the Causal-Generative Dual-Judge (CGDJ), auditing World Model Consistency from two perspectives. Explicit Causal Perception tests whether a generator reads a video scenario as a reasoning problem through spatio-temporal flattened visual question answering, while Implicit Generative Perception-Prediction Gap evaluates whether it renders the causal consequence as a consistent future video. Applying CGDJ to representative open- and closed-source generators reveals a clear Perception-Prediction Gap: open-source models produce plausible dynamics despite near-zero explicit causal perception, whereas advanced closed-source systems show stronger but still limited alignment between reasoning and generation. Further analysis exposes audio-visual misalignment, where models verbalize correct causal logic more reliably than they render it, challenging the "world simulator" narrative.

cs.CV↗

FedACT: Federated Adaptive Coordinate Trust Modulation for Robust Transformer Training under Data Heterogeneity

Federated Transformer training increasingly relies on local AdamW, whose adaptive updates can provide much stronger local progress than SGD-based training. However, under heterogeneous client data, even globally corrected AdamW updates may remain highly uneven in coordinate-wise reliability. We refer to this phenomenon as coordinate trust mismatch. Existing federated adaptive optimizers mainly address mismatch at the client-update or communication-round level, but still apply the corrected adaptive direction densely and uniformly across coordinates. In this paper, we propose FedACT, a global-aware coordinate trust modulation method for federated AdamW training. FedACT first forms a globally corrected adaptive direction and then reallocates update magnitudes according to a coordinate-wise trust score, assigning larger steps to coordinates jointly supported by local gradients and global correction, while preserving smaller non-zero updates on the remaining coordinates. Extensive experiments on federated vision Transformers, CNNs, LLM pre-training, and LLM fine-tuning show that FedACT consistently improves over strong federated adaptive baselines, with the largest gains on Transformer models under stronger data heterogeneity. Mechanism analyses further show that FedACT improves cross-client direction consistency, suggesting that coordinate-level trust allocation effectively complements round-level global-local correction. Code will be released.

cs.LG↗

EgoCogNav: Cognition-aware Human Egocentric Navigation

Modeling the cognitive and experiential factors of human navigation is central to deepening our understanding of human-environment interaction and to enabling safe social navigation and effective assistive wayfinding. Most existing methods focus on forecasting motions in fully observed scenes and often neglect human factors that capture how people feel and respond to space. To address this gap, we propose EgoCogNav, a multimodal egocentric navigation framework that jointly forecasts perceived path uncertainty, trajectories and head motion from egocentric video, gaze, and motion history. To facilitate research in the field, we introduce the Cognition-aware Egocentric Navigation (CEN) dataset consisting of 6 hours real-world egocentric recordings capturing diverse navigation behaviors in real-world scenarios. Experiments show that EgoCogNav learns the perceived uncertainty that strongly correlates with human-like behaviors such as scanning, hesitation, and backtracking while improving trajectory and head-motion forecasting on held-out navigation recordings.

cs.LG↗

Microwave Linear Analog Computer (MiLAC)-Aided MIMO Radar Sensing: Transmit Beamforming Design and DoA Estimation

Multiple-input multiple-output (MIMO) radar has waveform diversity and large spatial degrees of freedom (DoFs), making it attractive for high-resolution sensing. Scaling MIMO radar to massive arrays can further improve sensing performance, but it also increases hardware cost, power consumption, and digital processing complexity. The microwave linear analog computer (MiLAC) can tackle these challenges by moving linear operations from the digital domain to the analog domain. MiLAC has shown promising benefits for communications in recent studies and this paper identifies its potential for radar sensing. Specifically, we consider both MiLAC-aided transmit beamforming and receiver-side two-dimensional discrete Fourier transform (2D-DFT)-based direction-of-arrival (DoA) estimation. For transmit beamforming, we formulate a weighted Cramer Rao bound (CRB) minimization problem under lossless and reciprocal MiLAC constraints and propose a penalty dual decomposition (PDD)-based iterative algorithm to address the non-convex problem. We further prove that MiLAC-aided and fully-digital beamforming achieve the same CRB. For receiver processing, we show that the 2D DFT can be implemented by a lossless reciprocal MiLAC, which enables analog-domain DoA estimation without digital optimization. Numerical results confirm the theoretical finding and show that the MiLAC-aided approach achieves the same CRB and DoA estimation performance as the fully-digital benchmark. Meanwhile, hardware cost and power consumption are reduced because only low-resolution DACs are required at the transmitter, while RF chains and ADCs are eliminated at the receiver. Moreover, performing the 2D DFT in the analog domain eliminates all digital DFT operations for DoA estimation.

eess.SP↗

Enabling Full-Duplex LEO Satellite Systems with Non-Reciprocal BD-RIS-Assisted Beamforming

Low Earth orbit (LEO) satellites are a promising technology for providing low-latency, high-data-rate, and wide-coverage communication services. However, with growing demand for data transmission, future non-terrestrial networks (NTNs) require high spectral efficiency especially with low-gain antennas at the ground devices. This motivates the adoption of in-band full-duplex (FD) systems. In addition, the potential imbalance between downlink (DL) and uplink (UL) transmissions necessitates flexibility in resource allocation. To overcome these challenges, we propose an FD LEO satellite system, where the non-reciprocal beyond-diagonal reconfigurable intelligent surfaces (NR-BD-RIS) and multiple transmit and receive antennas are attached to the LEO satellite. NR-BD-RIS reflects the DL and UL signals by passive beamforming. By incorporating non-reciprocal components into the impedance network of RIS, the NR-BD-RIS breaks channel reciprocity, facilitating simultaneous support for multiple beam directions. To cover a wide coverage, we propose a time-sharing scheduling framework in which the NR-BD-RIS simultaneously serves multiple DL and multiple UL ground devices within each time slot. An optimization problem is defined to maximize the weighted sum-rate over the entire scheduling period. Numerical results demonstrate that the proposed NR-BD-RIS significantly performs better than both conventional BD-RIS and diagonal RIS (D-RIS) with respect to DL and UL sum-rate performance under both single-user (SU) and multiple-user (MU) cases. Additionally, NR-BD-RIS requires less frequent reconfiguration compared to the other two types of RIS, making it more practical for implementation.

eess.SP↗