TimeWarp: Evaluating Web Agents by Revisiting the Past
#TimeWarp #web agents #evaluation #benchmark #dynamic web #historical data #robustness #adaptability
📌 Key Takeaways
- TimeWarp introduces a new evaluation method for web agents by revisiting past web states.
- The approach tests agents' ability to handle dynamic web environments and historical data.
- It aims to improve robustness and adaptability in automated web interaction tasks.
- The method provides a benchmark for comparing performance across different web agent models.
📖 Full Retelling
arXiv:2603.04949v1 Announce Type: new
Abstract: The improvement of web agents on current benchmarks raises the question: Do today's agents perform just as well when the web changes? We introduce TimeWarp, a benchmark that emulates the evolving web using containerized environments that vary in UI, design, and layout. TimeWarp consists of three web environments, each with six UI versions spanning different eras of the internet, paired with a set of complex, realistic tasks requiring different for
🏷️ Themes
Web Agents, Evaluation
Entity Intersection Graph
No entity connections available yet for this article.
Original Source
--> Computer Science > Artificial Intelligence arXiv:2603.04949 [Submitted on 5 Mar 2026] Title: TimeWarp: Evaluating Web Agents by Revisiting the Past Authors: Md Farhan Ishmam , Kenneth Marino View a PDF of the paper titled TimeWarp: Evaluating Web Agents by Revisiting the Past, by Md Farhan Ishmam and Kenneth Marino View PDF Abstract: The improvement of web agents on current benchmarks raises the question: Do today's agents perform just as well when the web changes? We introduce TimeWarp, a benchmark that emulates the evolving web using containerized environments that vary in UI, design, and layout. TimeWarp consists of three web environments, each with six UI versions spanning different eras of the internet, paired with a set of complex, realistic tasks requiring different forms of web navigation. Our experiments reveal web agents' vulnerability to changes and the limitations of behavior cloning on single-version trajectories. To address this, we propose TimeTraj, a simple yet effective algorithm that uses plan distillation to collect trajectories across multiple versions. By training agents on teacher rollouts using our BC-variant, we achieve substantial performance gains: $20.4\%\rightarrow37.7\%$ for Qwen-3 4B and $0\%\rightarrow27.0\%$ for Llama-3.1 8B models. We hope our work helps researchers study generalization across web designs and unlock a new paradigm for collecting plans rather than trajectories, thereby improving the robustness of web agents. Subjects: Artificial Intelligence (cs.AI) ; Computation and Language (cs.CL); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG) Cite as: arXiv:2603.04949 [cs.AI] (or arXiv:2603.04949v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2603.04949 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Md Farhan Ishmam [ view email ] [v1] Thu, 5 Mar 2026 08:43:06 UTC (11,625 KB) Full-text links: Access Paper: View a PDF of the paper titled ...
Read full article at source