CowPilot: A Framework for Autonomous and Human-Agent Collaborative Web Navigation
Faria Huq , Zora Zhiruo Wang , Frank F. Xu , Tianyue Ou , Shuyan Zhou , Jeffrey P. Bigham , Graham Neubig
- 🏛 Institutions
- CMU
- 📅 Date
- April 30, 2025
- 📑 Publisher
- NAACL 2025 (System Demonstrations)
- 💻 Env
- Web
- 🔑 Keywords
TLDR
CowPilot is a mixed-initiative web-navigation framework where an agent proposes next steps while the user can pause, reject, override, or hand control back at any time. Across five websites, the collaborative mode reaches the highest success rate while requiring humans to perform only a small fraction of the total steps, and the system is also positioned as a data-collection and evaluation tool.
Related papers (24)
- OpeFlo: Automated UX Evaluation via Simulated Human Web Interaction with GUI GroundingFebruary 25, 2026 · arXiv
- ColorBrowserAgent: Complex Long-Horizon Browser Agent with Adaptive Knowledge EvolutionJanuary 12, 2026 · arXiv
- WebATLAS: An LLM Agent with Experience-Driven Memory and Action SimulationOctober 26, 2025 · NeurIPS 2025 Workshop on Language Agents and World Models
- Surfer 2: The Next Generation of Cross-Platform Computer Use AgentsOctober 22, 2025 · arXiv
- PolySkill: Learning Generalizable Skills Through Polymorphic AbstractionOctober 17, 2025 · arXiv
- Infogent: An Agent-Based Framework for Web Information AggregationApril 29, 2025 · Findings of NAACL 2025
- SkillWeaver: Web Agents can Self-Improve by Discovering and Honing SkillsApril 9, 2025 · arXiv
- Inducing Programmatic Skills for Agentic TasksApril 9, 2025 · COLM 2025
- LiteWebAgent: The Open-Source Suite for VLM-Based Web-Agent ApplicationsMarch 4, 2025 · NAACL 2025 System Demonstrations
- WorldGUI: An Interactive Benchmark for Desktop GUI Automation from Any Starting PointFebruary 12, 2025 · arXiv
- WebWalker: Benchmarking LLMs in Web TraversalJanuary 13, 2025 · arXiv
- Proposer-Agent-Evaluator (PAE): Autonomous Skill Discovery For Foundation Model Internet AgentsDecember 17, 2024 · ICML 2025 (Poster)
- The BrowserGym Ecosystem for Web Agent ResearchDecember 6, 2024 · TMLR
- AdaptAgent: Adapting Multimodal Web Agents with Few-Shot Learning from Human DemonstrationsNovember 24, 2024 · ACL 2025
- WebPilot: A Versatile and Autonomous Multi-Agent System for Web Task Execution with Strategic ExplorationAugust 28, 2024 · AAAI 2025
- Agent-E: From Autonomous Web Navigation to Foundational Design Principles in Agentic SystemsJuly 17, 2024 · arXiv
- OS-Copilot: Towards Generalist Computer Agents with Self-ImprovementFebruary 12, 2024 · LLMAgents @ ICLR 2024
- GPT-4V(ision) is a Generalist Web Agent, if GroundedJanuary 3, 2024 · ICML 2024
- SteP: Stacked LLM Policies for Web ActionsOctober 5, 2023 · COLM 2024
- LASER: LLM Agent with State-Space Exploration for Web NavigationSeptember 15, 2023 · arXiv
- A Real-World WebAgent with Planning, Long Context Understanding, and Program SynthesisJuly 24, 2023 · ICLR 2024 (Oral)
- Grounding Open-Domain Instructions to Automate Web Support TasksMarch 30, 2021 · NAACL 2021
- Reinforcement Learning on Web Interfaces Using Workflow-Guided ExplorationFebruary 24, 2018 · ICLR 2018 (Poster)
- MobileGym: A Verifiable and Highly Parallel Simulation Platform for Mobile GUI Agent ResearchMay 25, 2026 · arXiv