Language Models can Solve Computer Tasks

Studies whether language models can solve computer-control tasks through prompting alone, using Recursive Critique and Improve to iteratively refine action plans. On MiniWoB++, the approach outperforms prior LLM prompting baselines and competitive supervised or RL systems on several tasks.

• A Data-Driven Approach for Learning to Control Computers
  February 16, 2022 · ICML 2022
• Workflow-GYM: Towards Long-Horizon Evaluation of Computer-use Agentic tasks in Real-World Professional Fields
  June 9, 2026 · arXiv
• A11y-Compressor: A Framework for Enhancing the Efficiency of GUI Agent Observations through Visual Context Reconstruction and Redundancy Reduction
  May 1, 2026 · arXiv
• WindowsWorld: A Process-Centric Benchmark of Autonomous GUI Agents in Professional Cross-Application Environments
  April 30, 2026 · arXiv
• VLAA-GUI: Knowing When to Stop, Recover, and Search, A Modular Framework for GUI Automation
  April 23, 2026 · arXiv
• The Blind Spot of Agent Safety: How Benign User Instructions Expose Critical Vulnerabilities in Computer-Use Agents
  April 12, 2026 · arXiv
• HealthAdminBench: Evaluating Computer-Use Agents on Healthcare Administration Tasks
  April 10, 2026 · arXiv
• EE-MCP: Self-Evolving MCP-GUI Agents via Automated Environment Generation and Experience Learning
  April 10, 2026 · arXiv
• Preference Redirection via Attention Concentration: An Attack on Computer Use Agents
  April 9, 2026 · arXiv
• Gym-Anything: Turn any Software into an Agent Environment
  April 7, 2026 · arXiv
• IntentScore: Intent-Conditioned Action Evaluation for Computer-Use Agents
  April 6, 2026 · arXiv
• GPA: Learning GUI Process Automation from Demonstrations
  April 2, 2026 · arXiv
• HippoCamp: Benchmarking Contextual Agents on Personal Computers
  April 1, 2026 · arXiv
• GUIDE: Resolving Domain Bias in GUI Agents through Real-Time Web Video Retrieval and Plug-and-Play Annotation
  March 27, 2026 · arXiv
• CUA-Suite: Massive Human-annotated Video Demonstrations for Computer-Use Agents
  March 25, 2026 · arXiv
• CUAAudit: Meta-Evaluation of Vision-Language Models as Auditors of Autonomous Computer-Use Agents
  March 11, 2026 · HEAL @ CHI 2026 Workshop
• Video-Based Reward Modeling for Computer-Use Agents
  March 10, 2026 · arXiv
• SpecOps: A Fully Automated AI Agent Testing Framework in Real-World GUI Environments
  March 10, 2026 · ICSE 2026
• PIRA-Bench: A Transition from Reactive GUI Agents to GUI-based Proactive Intent Recommendation Agents
  March 9, 2026 · arXiv
• OSExpert: Computer-Use Agents Learning Professional Skills via Exploration
  March 9, 2026 · arXiv
• AgentOS: From Application Silos to a Natural Language-Driven Data Ecosystem
  March 9, 2026 · arXiv
• IntentCUA: Learning Intent-level Representations for Skill Abstraction and Multi-Agent Planning in Computer-Use Agents
  February 19, 2026 · AAMAS 2026
• Mobile-Agent-v3.5: Multi-platform Fundamental GUI Agents
  February 15, 2026 · arXiv
• When Benign Inputs Lead to Severe Harms: Eliciting Unsafe Unintended Behaviors of Computer-Use Agents
  February 9, 2026 · arXiv