XBOUND: Exploring the Capability Boundaries of Device-Control Agents through Trajectory Tree Exploration

• URL: http://arxiv.org/abs/2505.21279v1
• Date: Tue, 27 May 2025 14:49:30 GMT
• Title: XBOUND: Exploring the Capability Boundaries of Device-Control Agents through Trajectory Tree Exploration
• Abstract summary: This study introduces a new perspective on evaluation methods for Device-Control Agents (DC agents)<n>We propose the XBOUND evaluation method, which employs the calculation of a novel Explore Metric to delineate the capability boundaries of DC agents.<n>We evaluate the OS-Atlas and UI-TARS series, examining both the overall and specific performance across five common tasks.
• Score: 73.87038197602268
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Recent advancements in vision-language models (VLMs) have spurred increased interest in Device-Control Agents (DC agents), such as utilizing in-the-wild device control to manage graphical user interfaces. Conventional methods for assessing the capabilities of DC agents, such as computing step-wise action accuracy and overall task success rates, provide a macroscopic view of DC agents' performance; however, they fail to offer microscopic insights into potential errors that may occur in real-world applications. Conducting a finer-grained performance evaluation of DC agents presents significant challenges. This study introduces a new perspective on evaluation methods for DC agents by proposing the XBOUND evaluation method, which employs the calculation of a novel Explore Metric to delineate the capability boundaries of DC agents. Compared to previous evaluation methods, XBOUND focuses on individual states to assess the proficiency of DC agents in mastering these states. Furthermore, we have developed a ``pseudo'' episode tree dataset derived from Android Control test data. Utilizing this dataset and XBOUND, we comprehensively evaluate the OS-Atlas and UI-TARS series, examining both the overall and specific performance across five common tasks. Additionally, we select representative cases to highlight the current deficiencies and limitations inherent in both series. Code is available at https://github.com/sqzhang-lazy/XBOUND.

Related papers

• Open Evaluation Agent: Efficient and Promptable Evaluation of Visual Generative Models [74.1961755473754]
  Existing evaluation methods rely on rigid pipelines that overlook specific user needs and provide numerical results without clear explanations.<n>We propose the Evaluation Agent framework, which employs human-like strategies for efficient, dynamic, multi-round evaluations.<n>We show that Evaluation Agent reduces evaluation time to 10% of traditional methods while delivering comparable results.
  arXiv  Detail & Related papers  (2026-08-10T14:42:10Z)
• Agentic CLEAR: Automating Multi-Level Evaluation of LLM Agents [12.703661629160571]
  We present Agentic CLEAR, an automatic, dynamic, and easy-to-use evaluation framework.<n>It produces textual insights into the agent behavior on three levels of granularity: system, trace, and node.<n>Agentic CLEAR operates above the observability layer, enabling seamless integration and featuring an intuitive UI.
  arXiv  Detail & Related papers  (2026-05-21T15:26:02Z)
• Exploring Interaction Paradigms for LLM Agents in Scientific Visualization [7.22431217973039]
  This paper examines how different types of large language model (LLM) agents perform on scientific visualization (SciVis) tasks.<n>We compare three primary interaction paradigms, including domain-specific agents with structured tool use, computer-use agents, and general-purpose coding agents.
  arXiv  Detail & Related papers  (2026-04-30T15:22:28Z)
• AJ-Bench: Benchmarking Agent-as-a-Judge for Environment-Aware Evaluation [71.49152943451328]
  We introduce a benchmark AJ-Bench to evaluate Agent-as-a-Judge across three domains-search, data systems, and graphical user interfaces-comprising 155 tasks and 516 annotated trajectories.<n>Experiments demonstrate consistent performance gains over LLM-as-a-Judge baselines, while also revealing substantial open challenges in agent-based verification.
  arXiv  Detail & Related papers  (2026-04-20T13:23:38Z)
• Mini-BEHAVIOR-Gran: Revealing U-Shaped Effects of Instruction Granularity on Language-Guided Embodied Agents [58.64789157580261]
  Mini-BEHAVIOR-Gran is a new benchmark for controlled studies of instruction granularity in embodied AI.<n>We compare four candidate metrics for cross-task granularity quantification.<n>Using width to organize training and evaluation reveals a non-monotonic U-shaped relationship between instruction granularity and performance.
  arXiv  Detail & Related papers  (2026-04-18T14:57:51Z)
• GUIDE: Interpretable GUI Agent Evaluation via Hierarchical Diagnosis [17.279226216630065]
  We introduce GUIDE (GUI Understanding and Interpretable Diagnostic Evaluation), a framework that decomposes trajectory assessment into three sequential stages.<n>Subtask Diagnosis evaluates each unit in context, assigning a completion verdict and generating a structured error analysis with corrective recommendations.<n>We validate GUIDE on three benchmarks: an industrial e-commerce dataset of 932 trajectories, AGENTREWARDBENCH spanning five web agent tasks with 1302 trajectories, and AndroidBench for mobile device control.
  arXiv  Detail & Related papers  (2026-04-06T03:58:43Z)
• See, Plan, Snap: Evaluating Multimodal GUI Agents in Scratch [20.231957791642635]
  We introduce ScratchWorld, a benchmark for evaluating multimodal GUI agents on program-by-construction tasks in Scratch.<n>ScratchWorld comprises 83 curated tasks spanning four distinct problem categories: Create, Debug, Extend, and Compute.<n>To ensure reliable assessment, we propose an execution-based evaluation protocol that validates the functional correctness of the constructed Scratch programs.
  arXiv  Detail & Related papers  (2026-02-11T12:54:53Z)
• ANCHOR: Branch-Point Data Generation for GUI Agents [52.22377425487]
  End-to-end GUI agents for real desktop environments require large amounts of high-quality interaction data.<n>We present a trajectory expansion framework Anchor that bootstraps scalable desktop supervision from a small set of verified seed demonstrations.<n>Experiments on standard desktop benchmarks, OSWorld and WindowsAgentArena, show that models fine-tuned on our expanded corpus achieve consistent improvements.
  arXiv  Detail & Related papers  (2026-02-06T19:55:26Z)
• Agentic Reward Modeling: Verifying GUI Agent via Online Proactive Interaction [7.731207237810125]
  VAGEN is a framework that employs a verifier agent equipped with interaction tools to autonomously plan verification strategies.<n>We show that VAGEN significantly improves evaluation accuracy compared to LLM-as-a-Judge baselines.
  arXiv  Detail & Related papers  (2026-01-31T07:36:54Z)
• See, Think, Act: Teaching Multimodal Agents to Effectively Interact with GUI by Identifying Toggles [26.687510922403405]
  multimodal agents' inability to reliably execute toggle control instructions remains a key bottleneck.<n>We propose State-aware Reasoning (StaR), a training method that teaches agents to perceive the current toggle state, analyze the desired state from the instruction, and act accordingly.<n>Experiments on three multimodal agents demonstrate that StaR can improve toggle instruction execution accuracy by over 30%.
  arXiv  Detail & Related papers  (2025-09-17T01:14:14Z)
• Visual Document Understanding and Question Answering: A Multi-Agent Collaboration Framework with Test-Time Scaling [83.78874399606379]
  We propose MACT, a Multi-Agent Collaboration framework with Test-Time scaling.<n>It comprises four distinct small-scale agents, with clearly defined roles and effective collaboration.<n>It shows superior performance with a smaller parameter scale without sacrificing the ability of general and mathematical tasks.
  arXiv  Detail & Related papers  (2025-08-05T12:52:09Z)
• What Limits Virtual Agent Application? OmniBench: A Scalable Multi-Dimensional Benchmark for Essential Virtual Agent Capabilities [56.646832992178105]
  We introduce OmniBench, a cross-platform, graph-based benchmark with an automated pipeline for synthesizing tasks of controllable complexity.<n>We present OmniEval, a multidimensional evaluation framework that includes subtask-level evaluation, graph-based metrics, and comprehensive tests across 10 capabilities.<n>Our dataset contains 36k graph-structured tasks across 20 scenarios, achieving a 91% human acceptance rate.
  arXiv  Detail & Related papers  (2025-06-10T15:59:38Z)
• AgentCPM-GUI: Building Mobile-Use Agents with Reinforcement Fine-Tuning [82.42421823672954]
  AgentCPM-GUI is built for robust and efficient on-device GUI interaction.<n>Our training pipeline includes grounding-aware pre-training to enhance perception.<n>AgentCPM-GUI achieves state-of-the-art performance on five public benchmarks.
  arXiv  Detail & Related papers  (2025-06-02T07:30:29Z)
• BEDI: A Comprehensive Benchmark for Evaluating Embodied Agents on UAVs [3.239200269731315]
  Embodied Agents based on Unmanned Aerial Vehicles (UAVs) have shown significant potential in autonomous tasks.<n>Current evaluation methods for UAV-Embodied Agents (UAV-EAs) remain constrained by the lack of standardized benchmarks.<n>We propose BEDI (Benchmark for Embodied Drone Intelligence), a systematic and standardized benchmark designed for evaluating UAV-EAs.
  arXiv  Detail & Related papers  (2025-05-23T12:14:00Z)
• A Comprehensive Survey of Agents for Computer Use: Foundations, Challenges, and Future Directions [4.904229981437243]
  Agents for computer use (ACUs) are an emerging class of systems capable of executing complex tasks on digital devices.<n>Despite rapid progress, ACUs are not yet mature for everyday use.
  arXiv  Detail & Related papers  (2025-01-27T15:44:02Z)
• TestAgent: A Framework for Domain-Adaptive Evaluation of LLMs via Dynamic Benchmark Construction and Exploratory Interaction [29.72874725703848]
  Large language models (LLMs) are increasingly deployed to various vertical domains.<n>Current evaluation methods rely on static and resource-intensive datasets that are not aligned with real-world requirements.<n>We introduce two key concepts: textbfBenchmark+, which extends the traditional question-answer benchmark into a more flexible strategy-criterion'' format.<n>We propose textbftextscTestAgent, an agent-based evaluation framework that implements these concepts using retrieval-augmented generation and reinforcement learning.
  arXiv  Detail & Related papers  (2024-10-15T11:20:42Z)
• Unsupervised Model Diagnosis [49.36194740479798]
  This paper proposes Unsupervised Model Diagnosis (UMO) to produce semantic counterfactual explanations without any user guidance. Our approach identifies and visualizes changes in semantics, and then matches these changes to attributes from wide-ranging text sources.
  arXiv  Detail & Related papers  (2024-10-08T17:59:03Z)
• CableInspect-AD: An Expert-Annotated Anomaly Detection Dataset [14.246172794156987]
  $textitCableInspect-AD$ is a high-quality dataset created and annotated by domain experts from Hydro-Qu'ebec, a Canadian public utility. This dataset includes high-resolution images with challenging real-world anomalies, covering defects with varying severity levels. We present a comprehensive evaluation protocol based on cross-validation to assess models' performances.
  arXiv  Detail & Related papers  (2024-09-30T14:50:13Z)
• DMC-VB: A Benchmark for Representation Learning for Control with Visual Distractors [13.700885996266457]
  Learning from previously collected data via behavioral cloning or offline reinforcement learning (RL) is a powerful recipe for scaling generalist agents. We present theDeepMind Control Visual Benchmark (DMC-VB), a dataset collected in the DeepMind Control Suite to evaluate the robustness of offline RL agents. Accompanying our dataset, we propose three benchmarks to evaluate representation learning methods for pretraining, and carry out experiments on several recently proposed methods.
  arXiv  Detail & Related papers  (2024-09-26T23:07:01Z)
• Deep Learning for Video Anomaly Detection: A Review [52.74513211976795]
  Video anomaly detection (VAD) aims to discover behaviors or events deviating from the normality in videos. In the era of deep learning, a great variety of deep learning based methods are constantly emerging for the VAD task. This review covers the spectrum of five different categories, namely, semi-supervised, weakly supervised, fully supervised, unsupervised and open-set supervised VAD.
  arXiv  Detail & Related papers  (2024-09-09T07:31:16Z)
• DEAR: Disentangled Environment and Agent Representations for Reinforcement Learning without Reconstruction [4.813546138483559]
  Reinforcement Learning (RL) algorithms can learn robotic control tasks from visual observations, but they often require a large amount of data. In this paper, we explore how the agent's knowledge of its shape can improve the sample efficiency of visual RL methods. We propose a novel method, Disentangled Environment and Agent Representations, that uses the segmentation mask of the agent as supervision.
  arXiv  Detail & Related papers  (2024-06-30T09:15:21Z)
• Learning Feature Inversion for Multi-class Anomaly Detection under General-purpose COCO-AD Benchmark [101.23684938489413]
  Anomaly detection (AD) is often focused on detecting anomalies for industrial quality inspection and medical lesion examination. This work first constructs a large-scale and general-purpose COCO-AD dataset by extending COCO to the AD field. Inspired by the metrics in the segmentation field, we propose several more practical threshold-dependent AD-specific metrics.
  arXiv  Detail & Related papers  (2024-04-16T17:38:26Z)
• CCA: Collaborative Competitive Agents for Image Editing [55.500493143796405]
  This paper presents a novel generative model, Collaborative Competitive Agents (CCA)<n>It leverages the capabilities of multiple Large Language Models (LLMs) based agents to execute complex tasks.<n>The paper's main contributions include the introduction of a multi-agent-based generative model with controllable intermediate steps and iterative optimization.
  arXiv  Detail & Related papers  (2024-01-23T11:46:28Z)
• Diffusion-based Visual Counterfactual Explanations -- Towards Systematic Quantitative Evaluation [64.0476282000118]
  Latest methods for visual counterfactual explanations (VCE) harness the power of deep generative models to synthesize new examples of high-dimensional images of impressive quality. It is currently difficult to compare the performance of these VCE methods as the evaluation procedures largely vary and often boil down to visual inspection of individual examples and small scale user studies. We propose a framework for systematic, quantitative evaluation of the VCE methods and a minimal set of metrics to be used.
  arXiv  Detail & Related papers  (2023-08-11T12:22:37Z)
• RADDLE: An Evaluation Benchmark and Analysis Platform for Robust Task-oriented Dialog Systems [75.87418236410296]
  We introduce the RADDLE benchmark, a collection of corpora and tools for evaluating the performance of models across a diverse set of domains. RADDLE is designed to favor and encourage models with a strong generalization ability. We evaluate recent state-of-the-art systems based on pre-training and fine-tuning, and find that grounded pre-training on heterogeneous dialog corpora performs better than training a separate model per domain.
  arXiv  Detail & Related papers  (2020-12-29T08:58:49Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.