已审阅并标注「规划」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。
OpAgent combines multi-task GUI fine-tuning with online web reinforcement learning using outcome judgments and rule-based progress rewards. A modular planner, grounder, reflector and summarizer support navigation and recovery, evaluated separately from the underlying trained policy.
This survey organizes agent skills by architecture, acquisition, deployment and security, including their role in computer-use systems and GUI grounding. It also proposes a permission and lifecycle governance framework for skill deployment.
CATTS allocates additional inference computation when an agent’s action votes show uncertainty. The paper compares voting and arbiter strategies and evaluates this selective scaling on WebArena-Lite, Online-Mind2Web and GoBrowse.
WebTestPilot turns natural-language requirements into web test steps with inferred preconditions and postconditions over symbolized GUI elements. A benchmark of bug-injected web applications tests whether these context-aware oracles distinguish application defects from agent mistakes.
TreeCUA organizes GUI exploration trajectories as reusable trees and balances deeper tasks against broader coverage. Its data pipeline combines action verification, memory-backed recovery and a DPO variant that uses information from adjacent branches.
ANCHOR expands verified desktop demonstrations by generating new tasks at meaningful branch points and executing their instructions. State-aware verification and action filtering refine the resulting trajectories, which are evaluated through fine-tuned agents on OSWorld and WindowsAgentArena.
M2-Miner uses multi-agent-guided Monte Carlo tree search to collect mobile GUI intent–trajectory pairs. Intent recycling and progressive collector training aim to increase the diversity and usefulness of the mined interaction data.
Introduces MIRAGE, a web decision benchmark where agents must actively uncover evidence that can contradict the visible context. Separate investigation, reasoning, decision, and hallucination measures diagnose failures beyond reaching the right page.
Agent Alpha uses step-level Monte Carlo tree search to reuse successful prefixes and recover from early GUI mistakes. Comparison-based evaluation and diversity-constrained expansion guide the search, with experiments on OSWorld.
Avenir-Web combines multiple grounding specialists with procedural experience, a task checklist and adaptive memory. It evaluates this planning and execution framework on live Online-Mind2Web tasks.
SafePred uses world-model risk forecasts to guide action filtering and task replanning. Its desktop and web safety evaluations distinguish policy compliance, task success and success under the specified policy.
BEAP-Agent treats GUI execution as depth-first search with coordinated planning, execution and task-state tracking. It supports backtracking across multiple levels of earlier states and evaluates recovery on OSWorld.
CUA-Skill encodes Windows application knowledge in parameterized skills and execution/composition graphs. Its agent retrieves and instantiates these skills and uses memory for failure recovery, with end-to-end evaluation on WindowsAgentArena.
Uses execution uncertainty to trigger retrieval from application documentation, code, and past trajectories, organizing the results into structured AppCards for Android agent planning.
GraphPilot builds app-specific knowledge graphs from explored pages, elements and transitions, then uses them to plan action sequences. It requests additional interface observations when uncertain and checks plans against graph transition rules before correction and execution.
NOVA separates trusted task planning from screen-derived observations by producing a branching plan before execution and resolving runtime values through perception. Its OSWorld study examines security and utility while identifying remaining risks from perception-driven branch steering.
ExpSeek uses step-level entropy signals to decide when a web agent should request experience-based guidance. It supplies guidance tailored to the current step and evaluates the intervention across web-agent benchmarks.
OS-Symphony coordinates milestone-based reflection and memory with tools that build visually aligned tutorials through browser interaction. The framework evaluates trajectory correction and adaptation on online computer-use benchmarks including OSWorld.
MobileDreamer predicts post-action GUI states as task-relevant textual sketches with an order-invariant learning objective. Agents use these predictions for lookahead action selection, evaluated on AndroidWorld.
NestBrowse organizes browser actions in a nested framework that separates interaction control from page exploration. It evaluates this design for deep information seeking beyond snippet retrieval and direct URL fetching.
Evaluates semantic world models for mobile interfaces and studies natural-language state transitions as a tool for agent planning.
Introduces GAIR, which combines information from specialized GUI models through a general-purpose coordinating model. When the available information is insufficient, group reflection sends targeted requests back to the specialists before selecting an operation.
This work trains high-level schedulers with staged reinforcement learning driven by execution feedback. Its CES framework separates a Coordinator for planning and decomposition, a StateTracker for compressed context, and a low-level Executor.
Co-EPG alternates reinforcement learning for a planner with updates to a grounding model using generated training examples. The evolving grounder provides feedback to the planner, and evaluation uses Multimodal Mind2Web and AndroidControl.
This work retrieves online video tutorials and converts them into structured GUI action demonstrations with short objectives. A two-stage selection process supplies a relevant local trajectory as in-context guidance at each execution step.
Combines structured progress-and-decision reasoning with history summarization to improve cross-domain generalization in GUI navigation agents.
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.
Surfer2 combines hierarchical context management with separate planning and execution, self-verification, and recovery for visual computer use. It evaluates the framework across web, desktop, and Android tasks without task-specific fine-tuning.
Introduces Hi-Agent, jointly training a high-level reasoning model and a low-level mobile action model. Execution feedback supplies a foresight advantage for optimizing subgoal decisions, with evaluations on Android control benchmarks.
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.
Guides UI agents with symbolic planning over a UI transition graph initialized from app-code analysis and updated through online exploration, with evaluation on AndroidWorld.
Introduces a Declarative Model Interface that maps existing GUIs to access, state and observation primitives while handling low-level interactions. Evaluation uses Word, PowerPoint and Excel on Windows with a GUI-agent baseline.
GUI-Spotlight trains a grounding model to reason over images and call tools that iteratively narrow its visual focus. The paper evaluates this refinement method on ScreenSpot-Pro after training on 18,500 samples.
PAL-UI combines observation/action summaries with selective retrieval of historical screenshots during planning. It trains mobile navigation models from step-level supervision and evaluates transfer to web navigation without additional training.
RAG-GUI trains a small guidance adapter to filter and summarize retrieved web tutorials while keeping the acting agent fixed. Supervised and rejection-sampling stages prepare the adapter for inference-time assistance, tested on offline web/mobile tasks and interactive Android tasks.
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.
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.
GUI-ReWalk begins with stochastic interface exploration and then infers goals to generate coherent action trajectories. It supports long-horizon workflows across applications and studies training on the resulting data using grounding and navigation benchmarks.
BTL-UI separates GUI interaction into screen-region attention, planning and executable-command generation. It adds an automated annotation pipeline and rule-based process/outcome rewards for training, with evaluation on both static understanding and dynamic interaction.
InfraMind explores industrial management GUIs using virtual-machine snapshots and builds memory for planning and hierarchical state recognition. Its framework also studies knowledge distillation and safety checks on open-source and commercial data-center management platforms.
Describes a mobile-agent stack combining region-based coordinate correction, hierarchical task planning, multi-agent voting, and reuse of successful interaction traces.
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.
UItron combines GUI perception and planning data engineering with supervised training and curriculum reinforcement learning in connected mobile and PC environments. It also collects more than one million operation steps across 100 Chinese apps and builds offline and online evaluation environments.
CODA combines a generalist planner with a specialist GUI executor and trains application-specific planners through decoupled GRPO. Successful specialist trajectories then support supervised training of a consolidated planner, evaluated across four ScienceBoard applications.
Proposes a proactive GUI assistant that gathers information across domains in response to user requests and anticipated information needs.
Introduces PerPilot and the PerInstruct dataset for completing user-specific mobile instructions through memory retrieval, app exploration, and clarification when missing context remains unresolved.
WebSight uses a vision-language model trained on web-interface examples within a modular agent with planning, action, verification, and episodic memory components. It evaluates element clicking on Showdown Clicks and full browsing tasks on WebVoyager using visual observations.
Introduces GUI-Owl models and Mobile-Agent-v3 orchestration, using environment-generated trajectories to support GUI capabilities.
Presents demonstration collection infrastructure, the AgentNet dataset, and a training pipeline for open computer-use models.
Introduces CogniWeb, which switches between fast reactive processing and slower deliberative reasoning according to task complexity. WebArena experiments examine this combination of learned behaviors and online planning.
The Self-Evolution Agent combines automatic generation of verifiable computer-use trajectories with step-wise reinforcement learning for long-horizon training. It also describes a model-enhancement procedure that integrates grounding and planning capabilities without an additional training stage.
CoAct-1 delegates subtasks between a GUI operator and a programmer that executes Python or Bash. Its OSWorld evaluation studies a hybrid action space in which scripts handle suitable file and data operations while visual interaction handles other steps.
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.
Builds reusable page memories from past mobile trajectories and retrieves them to support task planning and execution.
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 proactive assistance in visual-analytics interfaces, using interaction logs to detect support needs and provide context-aware suggestions and guidance.
GTA1 samples multiple action proposals at each GUI step and uses a judge to select among them. A separate grounding model is trained with reinforcement learning to map selected proposals to visual targets, linking planning-time search with click localization.
第 61–120 条，共 283 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。
