已审阅并标注「评测」的研究与资产。
条目依据已审阅标签归类，同一主题不表示资源之间存在依赖关系。日期保留来源精度，注释保留原语言。
OSWorld-MCP evaluates how agents choose between GUI operations and MCP tool calls in computer tasks. It combines generated and curated tools, with 158 manually validated tools across seven applications, to examine tool use alongside task completion.
Introduces MobileRisk-Live and OS-Sentinel, combining a formal verifier for system-level violations with a multimodal judge for contextual risks in mobile GUI workflows.
Introduces WebGraphEval, which merges multiple agents' trajectories into a weighted graph of canonical actions. Structural analyses examine recurring behavior, inefficient paths and important decision points beyond a final success score.
AndroidControl-Curated audits ambiguities and factual errors in AndroidControl and produces a refined evaluation dataset. The paper also studies post-training a small Magma-R1 model on a curated subset to examine the effect of data quality on GUI action assessment.
Presents LiteAgent for recording web-agent interactions and TrickyArena for enabling or disabling dark patterns in controlled websites. Experiments compare how individual patterns, combined patterns and visual or HTML changes affect agent decisions.
ColorBench represents mobile tasks as graphs of states collected from real devices, providing a quasi-dynamic environment with multiple valid paths and error paths. It evaluates single-app and cross-app tasks using completion, subtask progress, and atomic diagnostics.
Introduces HackWorld for evaluating visual computer-use agents in a capture-the-flag collection of web applications. The benchmark examines interface navigation and multistep planning during vulnerability discovery and exploitation tasks.
R-WoM evaluates how well world models predict nearby GUI states, align complete procedures, and recognize milestone transitions. It retrieves external tutorials to guide simulated future interactions, with experiments on subsets of OSWorld and WebArena.
Introduces SusBench, an online benchmark that inserts dark-pattern designs into consumer websites for agent evaluation. A human study checks the realism of the interventions, and agent trials compare susceptibility across pattern types.
Introduces SecureWebArena with simulated interactive websites, an attack-vector taxonomy and labeled agent trajectories. Its evaluation separates failures in reasoning, behavioral traces and task outcomes across vision-language agent families.
WARC-Bench packages dynamic web pages as replayable archives and evaluates short navigation subtasks involving components such as date pickers and scrollable containers. It also studies supervised training followed by reinforcement learning with verifiable rewards on these browser interactions.
Introduces BrowserArena, a live-web platform for head-to-head agent comparisons on user-submitted tasks. Step-level human annotations support targeted analysis of navigation, pop-up handling and CAPTCHA-related failures.
Introduces Behavior Judge, which converts execution trajectories into behavior narratives and compares multiple candidate rollouts. Experiments examine this selection approach on OSWorld, WindowsAgentArena and AndroidWorld.
BLIND-ACT tests whether computer-use agents persist with ambiguous, contradictory, or infeasible goals without adequate contextual reasoning. Its OSWorld-based tasks and judged trajectories examine these behaviors and the effect of mitigation prompts.
Introduces WAInjectBench with a threat-model taxonomy and malicious and benign text-image datasets for web-agent prompt-injection detection. It compares detector families across explicit and less visible forms of injected content.
SCUBA evaluates enterprise UI navigation and CRM workflows with 300 task instances derived from Salesforce user interviews. Salesforce sandbox execution and milestone-based assessment support comparisons between zero-shot agents and agents given demonstrations.
GMS pairs a general vision-language scanner that proposes interface regions with a specialist locator that predicts coordinates within them. Hierarchical coarse-to-fine search coordinates the two components for GUI grounding on ScreenSpot-Pro.
WAREX augments existing web-agent benchmarks with reliability challenges arising from unstable systems and changing or disruptive website behavior. Evaluation on WebArena, WebVoyager and REAL examines how these disturbances affect task completion.
GUI-PRA evaluates GUI progress by generating task-specific criteria and actively inspecting visual evidence with tools. It tests offline judgment on OS-Critic Bench and the effects of this feedback on online mobile policies in AndroidWorld.
RISK combines browser-interaction training data, a benchmark, and reinforcement fine-tuning for e-commerce risk-management workflows. Its trajectories and rewards distinguish individual actions from multi-step stateful tasks involving deeply embedded web content.
ProRe evaluates GUI task outcomes by having a reasoner assign targeted environment-probing tasks to specialist actors. The additional observations support reward judgments, with separate evaluation of judge accuracy and downstream agent task success.
Evaluates multimodal LLMs on web interface reasoning, visual perturbation robustness, and safety-critical web understanding tasks.
Introduces a modular framework for identifying errors at intermediate stages of a web-agent pipeline. A SeeAct and Mind2Web case study examines weaknesses that are obscured by final task-success metrics.
Introduces a benchmark for binary GUI-toggle instructions and State-aware Reasoning, which compares the observed control state with the requested state before choosing an action. It evaluates agents on toggle tasks and broader interactive benchmarks.
UI-S1 simulates multi-turn online learning over offline trajectories, using a patch module to recover divergences while retaining the model’s own outputs in context. It combines discounted future returns with step- and episode-level advantages and introduces a semi-online performance proxy.
Introduces Chameleon, an environmental-injection evaluation framework that varies webpage content and trigger context. It combines simulated page variation with attention-based optimization and tests GUI agents across realistic websites.
Studies how GUI agents, people, and human-agent teams respond to manipulative interface designs. The two-phase evaluation examines failure modes and the benefits and costs of human oversight.
Introduces AgentSentinel, which intercepts sensitive computer-agent operations and audits them using task context together with system traces. BadComputerUse evaluates this enforcement approach across attack scenarios involving tool and computer control.
MAS-Bench evaluates mobile agents that combine GUI actions with APIs, deep links, and reusable scripts. Its 139 tasks across 11 apps test use of 88 predefined shortcuts as well as the ability to generate new reusable workflows.
Introduces SparkUI-Parser, which adds continuous coordinate decoding and a rejection mechanism for nonexistent targets to a multimodal model. It also presents ScreenParse for evaluating structural interface perception.
MobileRAG combines retrieval and memory components to support mobile tasks that need external knowledge or long action sequences. MobileRAG-Eval is introduced to study these requirements on complex real-world app tasks.
Introduces reasoning gates that require agents to solve generated, verifiable puzzles before receiving access to resources. The throttling approach is deployed on a custom website and MCP servers and evaluated with web agents.
FlashAdventure evaluates GUI agents on completing story arcs in 34 adventure games. It pairs an automated gameplay judge with COAST, a framework that keeps long-term clue memory to plan and complete sequential objectives.
KG-RAG converts UI transition graphs into a searchable vector store and uses intent-guided retrieval to construct actionable navigation paths. It introduces KG-Android-Bench and KG-Harmony-Bench and evaluates transfer from mobile apps to web and desktop tasks.
Introduces Morae, a UI agent that detects decision points and pauses for user choices using interface code and screenshots. A study with blind and low-vision participants evaluates this mixed-initiative approach on real web tasks.
InquireMobile trains mobile agents to seek human confirmation at critical decisions using staged reinforcement fine-tuning and pre-action reasoning. InquireBench evaluates safe interaction and proactive inquiry, distinguishing asking for help from completing the underlying task.
Introduces SAPA-Bench with 7,138 smartphone scenarios to evaluate whether MLLM-powered mobile agents detect and protect sensitive personal information on screen.
Amazon-Bench generates functionality-specific web tasks from page content and interactive controls, covering operations beyond product search. Its evaluator checks task completion and unintended changes to account state, such as address or purchase settings.
WebMall provides four simulated online shops with heterogeneous product data for comparison shopping, related-product searches, and checkout tasks. It evaluates web agents with different observation formats and memory configurations in a fixed offline environment.
RealDevWorld evaluates generated software repositories with 194 development tasks and AppEvalPilot, an agent judge that operates the resulting GUIs. The evaluation examines functional behavior, visual fidelity, and runtime responses to produce task-specific diagnostic feedback.
FineState-Bench evaluates fine-grained GUI perception and control across desktop, web, and mobile interfaces. Its Visual Diagnostic Assistant separates perception and localization contributions in controlled interaction experiments.
MVISU-Bench evaluates bilingual mobile tasks involving cross-app work, vague requests, user interaction and unethical instructions alongside ordinary single-app tasks. Its Aider prompting module is designed to clarify intent and mitigate risky execution within agent workflows.
RecAgent narrows attention to recommended UI components and requests user feedback when task intent or decisions remain uncertain. It introduces ComplexAction to evaluate specified single-step GUI actions in difficult interface scenarios.
Web-CogReasoner separates factual and conceptual web knowledge from procedural reasoning, using a structured curriculum to train a multimodal agent. Web-CogDataset and Web-CogBench support study of these capabilities alongside web-navigation evaluation.
NaturalGAIA builds verifiable long-horizon GUI tasks from human interaction intents while separating causal task structure from its verbal description. Its LightManus-Jarvis framework divides dynamic planning and context management from execution based on visual and structural interface information.
MMBench-GUI evaluates interface understanding, element grounding, task automation, and collaboration across six named desktop, mobile, and web platforms. It introduces the Efficiency-Quality Area metric for examining execution efficiency alongside task quality in online GUI automation.
OS-MAP organizes 416 computer-use tasks across 15 applications by a five-level automation taxonomy and a hierarchy of user demands. The benchmark separates required autonomy from generalization scope to diagnose where agents struggle as task demands increase.
Studies indirect prompt injection through HTML content consumed as accessibility trees. The evaluation tests how manipulated page observations redirect browser-agent actions on real websites.
Introduces WebGuard, a human-annotated dataset for predicting outcomes and risk levels of state-changing website actions. It provides generalization splits and evaluates specialized guardrail models.
Compares DOM-based and pixel-based web agents on a cloned news site containing advertisements, consent dialogs and paywalls. The study examines how interface presentation affects ad interaction and consequential subscription choices.
Studies backdoors that redirect GUI visual grounding even when the task plan is correct. VisualTrap evaluates whether poisoned grounding training transfers to downstream tasks and across web, mobile, and desktop interface settings.
Evaluates agent safety across real browser, code, filesystem, shell, and messaging tools using rule-based and model-based assessments.
Analyzes security boundaries in computer-use agents through a risk taxonomy and concrete adversarial case studies. It links observed failures to input provenance, interface-action binding, memory, and delegation, and proposes CUA-specific evaluation principles.
Introduces an app-content instrumentation framework and complementary interactive and static test suites for mobile GUI agents. It studies how untrusted third-party content changes task execution and state understanding.
Studies compute allocation for web-agent post-training with teacher-generated supervised fine-tuning followed by on-policy reinforcement learning. It samples 1,370 training configurations and uses bootstrap analysis to examine hyperparameter sensitivity on WorkArena and MiniWoB++.
WebArxiv provides 510 static-snapshot tasks with deterministic answers for scholarly browsing, including constrained paper retrieval, content extraction, and cross-paper comparison. The study also tests dynamic retrieval of interaction history to address repetitive or incomplete agent reasoning.
Studies chain-level risk detection for mobile multimodal agents and a language-model-assisted evaluator, assessing safety interventions together with functional preservation.
CogniGUI combines hierarchical visual parsing of interface elements with a GRPO-based grounding agent that compares alternative interaction paths. Its ScreenSeek benchmark evaluates navigation across applications, changing interface states, and consistency between interfaces.
OSWorld-Human adds human-determined trajectories to OSWorld to evaluate the efficiency of desktop computer-use agents. The study compares 16 agents and analyzes model-call latency, changes in step duration during task execution, and excess action steps.
Studies context manipulation in web agents by altering stored task representations rather than webpage prompts alone. Evaluation on Browser-use and Agent-E examines memory-directed attacks and the implications for secure context handling.
第 361–420 条，共 723 条资源。
通过主题导出清单获取完整记录和证据。克隆目录后，可将主题记录直接交给标准 RAG 与引用导出工具。
