Searcharxiv⌕ Search

arXiv subjects

Zilu Zhang

Publications and source records attributed to Zilu Zhang.

10 recordsLinked to original sources
arXiv2026

Resource Consumption Threats in Large Language Models

Given limited and costly computational infrastructure, resource efficiency is a key requirement for large language models (LLMs). Efficient LLMs increase service capacity for providers and reduce latency and API costs for users. Recent resource consumption threats induce excessive generation, degrading model efficiency and harming both service availability and economic sustainability. This survey presents a systematic review of threats to resource consumption in LLMs. We further establish a unified view of this emerging area by clarifying its scope and examining the problem along the full pipeline from threat induction to mechanism understanding and mitigation. Our goal is to clarify the problem landscape for this emerging area, thereby providing a clearer foundation for characterization and mitigation.

cs.CR↗
arXiv2026

Ego to World: Collaborative Spatial Reasoning in Embodied Systems via Reinforcement Learning

Understanding the world from distributed, partial viewpoints is a fundamental challenge for embodied multi-agent systems. Each agent perceives the environment through an ego-centric view that is often limited by occlusion and ambiguity. To study this problem, we introduce the Ego-to-World (E2W) benchmark, which evaluates a vision-language model's ability to fuse heterogeneous viewpoints across three tasks: (i) global counting, (ii) relational location reasoning, and (iii) action-oriented grasping that requires predicting view-specific image coordinates. To address this setting, we propose CoRL, a two-stage framework that combines Chain-of-Thought supervised fine-tuning with reinforcement learning using Group-Relative Policy Optimization. Its core component, the Cross-View Spatial Reward (CVSR), provides dense task-aligned feedback by linking reasoning steps to visual evidence, ensuring coherent cross-view entity resolution, and guiding the model toward correct final predictions. Experiments on E2W show that CoRL consistently surpasses strong proprietary and open-source baselines on both reasoning and perception-grounding metrics, while ablations further confirm the necessity of each CVSR component. Beyond that, CoRL generalizes to external spatial reasoning benchmarks and enables effective real-world multi-robot manipulation with calibrated multi-camera rigs, demonstrating cross-view localization and successful grasp-and-place execution. Together, E2W and CoRL provide a principled foundation for learning world-centric scene understanding from distributed, ego-centric observations, advancing collaborative embodied AI.

cs.RO↗
arXiv2026

HomeSafe-Bench: Evaluating Vision-Language Models on Unsafe Action Detection for Embodied Agents in Household Scenarios

The rapid evolution of embodied agents has accelerated the deployment of household robots in real-world environments. However, unlike structured industrial settings, household spaces introduce unpredictable safety risks, where system limitations such as perception latency and lack of common sense knowledge can lead to dangerous errors. Current safety evaluations, often restricted to static images, text, or general hazards, fail to adequately benchmark dynamic unsafe action detection in these specific contexts. To bridge this gap, we introduce HomeSafe-Bench, a challenging benchmark designed to evaluate Vision-Language Models (VLMs) on unsafe action detection in household scenarios. HomeSafe-Bench is contrusted via a hybrid pipeline combining physical simulation with advanced video generation and features 438 diverse cases across six functional areas with fine-grained multidimensional annotations. Beyond benchmarking, we propose Hierarchical Dual-Brain Guard for Household Safety (HD-Guard), a hierarchical streaming architecture for real-time safety monitoring. HD-Guard coordinates a lightweight FastBrain for continuous high-frequency screening with an asynchronous large-scale SlowBrain for deep multimodal reasoning, effectively balancing inference efficiency with detection accuracy. Evaluations demonstrate that HD-Guard achieves a superior trade-off between latency and performance, while our analysis identifies critical bottlenecks in current VLM-based safety detection.

cs.CV↗
arXiv2025

Text2Traffic: A Text-to-Image Generation and Editing Method for Traffic Scenes

With the rapid advancement of intelligent transportation systems, text-driven image generation and editing techniques have demonstrated significant potential in providing rich, controllable visual scene data for applications such as traffic monitoring and autonomous driving. However, several challenges remain, including insufficient semantic richness of generated traffic elements, limited camera viewpoints, low visual fidelity of synthesized images, and poor alignment between textual descriptions and generated content. To address these issues, we propose a unified text-driven framework for both image generation and editing, leveraging a controllable mask mechanism to seamlessly integrate the two tasks. Furthermore, we incorporate both vehicle-side and roadside multi-view data to enhance the geometric diversity of traffic scenes. Our training strategy follows a two-stage paradigm: first, we perform conceptual learning using large-scale coarse-grained text-image data; then, we fine-tune with fine-grained descriptive data to enhance text-image alignment and detail quality. Additionally, we introduce a mask-region-weighted loss that dynamically emphasizes small yet critical regions during training, thereby substantially enhancing the generation fidelity of small-scale traffic elements. Extensive experiments demonstrate that our method achieves leading performance in text-based image generation and editing within traffic scenes.

cs.CV↗
arXiv2025

InternVL-X: Advancing and Accelerating InternVL Series with Efficient Visual Token Compression

Most multimodal large language models (MLLMs) treat visual tokens as "a sequence of text", integrating them with text tokens into a large language model (LLM). However, a great quantity of visual tokens significantly increases the demand for computational resources and time. In this paper, we propose InternVL-X, which outperforms the InternVL model in both performance and efficiency by incorporating three visual token compression methods. First, we propose a novel vision-language projector, PVTC. This component integrates adjacent visual embeddings to form a local query and utilizes the transformed CLS token as a global query, then performs point-to-region cross-attention through these local and global queries to more effectively convert visual features. Second, we present a layer-wise visual token compression module, LVTC, which compresses tokens in the LLM shallow layers and then expands them through upsampling and residual connections in the deeper layers. This significantly enhances the model computational efficiency. Futhermore, we propose an efficient high resolution slicing method, RVTC, which dynamically adjusts the number of visual tokens based on image area or length filtering. RVTC greatly enhances training efficiency with only a slight reduction in performance. By utilizing 20% or fewer visual tokens, InternVL-X achieves state-of-the-art performance on 7 public MLLM benchmarks, and improves the average metric by 2.34% across 12 tasks.

cs.CV↗
arXiv2019

gfoRmula: An R package for estimating effects of general time-varying treatment interventions via the parametric g-formula

Researchers are often interested in using longitudinal data to estimate the causal effects of hypothetical time-varying treatment interventions on the mean or risk of a future outcome. Standard regression/conditioning methods for confounding control generally fail to recover causal effects when time-varying confounders are themselves affected by past treatment. In such settings, estimators derived from Robins's g-formula may recover time-varying treatment effects provided sufficient covariates are measured to control confounding by unmeasured risk factors. The package gfoRmula implements in R one such estimator: the parametric g-formula. This estimator easily adapts to binary or continuous time-varying treatments as well as contrasts defined by static or dynamic, deterministic or random treatment interventions, as well as interventions that depend on the natural value of treatment. The package accommodates survival outcomes as well as binary or continuous end of follow-up outcomes. For survival outcomes, the package has different options for handling competing events. This paper describes the gfoRmula package, along with motivating background, features, and examples.

stat.CO↗
arXiv2018

A distributed regression analysis application based on SAS software. Part I: Linear and logistic regression

Previous work has demonstrated the feasibility and value of conducting distributed regression analysis (DRA), a privacy-protecting analytic method that performs multivariable-adjusted regression analysis with only summary-level information from participating sites. To our knowledge, there are no DRA applications in SAS, the statistical software used by several large national distributed data networks (DDNs), including the Sentinel System and PCORnet. SAS/IML is available to perform the required matrix computations for DRA in the SAS system. However, not all data partners in these large DDNs have access to SAS/IML, which is licensed separately. In this first article of a two-paper series, we describe a DRA application developed for use in Base SAS and SAS/STAT modules for linear and logistic DRA within horizontally partitioned DDNs and its successful tests.

stat.CO↗
arXiv2018

A distributed regression analysis application based on SAS software Part II: Cox proportional hazards regression

Previous work has demonstrated the feasibility and value of conducting distributed regression analysis (DRA), a privacy-protecting analytic method that performs multivariable-adjusted regression analysis with only summary-level information from participating sites. To our knowledge, there are no DRA applications in SAS, the statistical software used by several large national distributed data networks (DDNs), including the Sentinel System and PCORnet. SAS/IML is available to perform the required matrix computations for DRA in the SAS system. However, not all data partners in these large DDNs have access to SAS/IML, which is licensed separately. In this second article of a two-paper series, we describe a DRA application developed using Base SAS and SAS/STAT modules for distributed Cox proportional hazards regression within horizontally partitioned DDNs and its successful tests.

stat.CO↗
arXiv2018

A Semantic Loss Function for Deep Learning with Symbolic Knowledge

This paper develops a novel methodology for using symbolic knowledge in deep learning. From first principles, we derive a semantic loss function that bridges between neural output vectors and logical constraints. This loss function captures how close the neural network is to satisfying the constraints on its output. An experimental evaluation shows that it effectively guides the learner to achieve (near-)state-of-the-art results on semi-supervised multi-class classification. Moreover, it significantly increases the ability of the neural network to predict structured objects, such as rankings and paths. These discrete concepts are tremendously difficult to learn, and benefit from a tight integration of deep learning and symbolic reasoning methods.

cs.AI↗