Elamur Research⌕
☰
home / papers / 2026 / SynthAgent: Adapting Web Agents with Synthetic Supervision

SynthAgent: Adapting Web Agents with Synthetic Supervision

Paper · 2026 · Zhaoyang Wang, Yiming Liang, Xuchao Zhang et al.

Summary

Web agents often fail on websites they never saw during training. Such sites rarely come with tasks or demonstrations, and hiring people to write them for each new site is costly. Earlier synthetic-data pipelines fix only one half of the task–trajectory pair. Some produce hallucinated, infeasible goals. Others rewrite goals mid-execution, which leaves early actions out of step with the final task. SynthAgent supervises both sides with fully synthetic data and no human labels. It has four stages:

  1. Task generation: tasks are proposed by exploring web elements grouped into functional categories.
  2. Task refinement: a task is revised during collection only when it clearly conflicts with what the agent observes.
  3. Trajectory refinement: each finished trajectory is cleaned using its full context to remove redundant or misaligned steps.
  4. Fine-tuning: an open-source agent is trained on the result with standard SFT.

GPT-4.1 drives the synthesis. Evaluated on WebArena and Online-Mind2Web, SynthAgent beats prior synthetic-data methods. This shows that adapting agents to new environments depends on data quality, not only quantity. The code is released publicly.

Key points

  • Ablations: the components appear to work together, since dropping all of them hurts overall.
    • Without task refinement, performance falls by 4.87%, the largest drop.
    • Without trajectory refinement, it falls by 3.99%.
    • Without categorized exploration, it falls by 3.54%, including −7.01% on CMS.
  • Efficiency: after trajectory refinement, 96.5% of tasks reach completion, at about 60% of Explorer's OpenAI API cost.
  • Scaling: going from 20 to 500 synthetic tasks per website raises average success from roughly 10.6 to 20.8.
    • Map plateaus at around 40% of the data.
    • CMS and Gitlab keep improving as data grows.
  • Benchmarks: WebArena is offline and self-hosted. Online-Mind2Web covers 136 live websites and tests generalization.
  • Limitations: the authors used only GPT-4.1 for synthesis and only plain SFT for training. They leave DPO and online RL for future work.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

 Institutions

🏢MicrosoftRedmond · United States325 works · 142,931 citations🎓University of North Carolina at Chapel HillChapel Hill · United States67 works · 16,474 citations🎓Purdue University West LafayetteWest Lafayette · United States23 works · 1,417 citations

 Fields, topics and concepts

📚 Shelves 5The field of science this work belongs to (OpenAlex fields).Computer Science · 16,667 worksArtificial Intelligence · 16,194 worksArts and Humanities · 16,039 worksLanguage and Linguistics · 15,891 worksInformation Systems · 1,437 works

Cite

Cite as Zhaoyang Wang, Yiming Liang, Xuchao Zhang, Qianhui Wu, Siwei Han, Anson Bastos, Rujia Wang, Chetan Bansal, Baolin Peng, Jianfeng Gao, Saravan Rajmohan and Huaxiu Yao. 2026. SynthAgent: Adapting Web Agents with Synthetic Supervision. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers).

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2026.acl-long.716

@inproceedings{wang2026synthagent,
  title = {{SynthAgent: Adapting Web Agents with Synthetic Supervision}},
  author = {Zhaoyang Wang, Yiming Liang, Xuchao Zhang, Qianhui Wu, Siwei Han, Anson Bastos, Rujia Wang, Chetan Bansal, Baolin Peng, Jianfeng Gao, Saravan Rajmohan and Huaxiu Yao},
  booktitle = {Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)},
  year = {2026},
  doi = {10.18653/v1/2026.acl-long.716},
  url = {https://doi.org/10.18653/v1/2026.acl-long.716}
}

References

51 references, 25 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Saaket Agashe, Kyle Wong, Vincent Tu, Jiachen Yang, Ang Li, and Xin Eric Wang. 2025. Agent s2: A compositional generalist-specialist framework for computer use agents. arXiv preprint arXiv:2504.00906. — [arXiv:2504.00906](https://arxiv.org/abs/2504.00906) find
  2. Shuai Bai, Yuxuan Cai, Ruizhe Chen, Keqin Chen, Xionghui Chen, Zesen Cheng, Lianghao Deng, Wei Ding, Chang Gao, Chunjiang Ge, Wenbin Ge, Zhifang Guo, Qidong Huang, and 1 others. 2025. Qwen3-vl technical report. Preprint, arXiv:2511.21631.
  3. Dongping Chen, Yue Huang, Siyuan Wu, Jingyu Tang, Liuyi Chen, Yilin Bai, Zhigang He, Chenlong Wang, Huichi Zhou, Yiqiang Li, and 1 others. 2024a. Gui-world: A dataset for gui-oriented multimodal llm-based agents. arXiv e-prints, pages arXiv–2406. find
  4. Zehui Chen, Kuikun Liu, Qiuchen Wang, Wenwei Zhang, Jiangning Liu, Dahua Lin, Kai Chen, and Feng Zhao. 2024b. Agent-flan: Designing data and methods of effective agent tuning for large language models. arXiv preprint arXiv:2403.12881.
  5. Xiang Deng, Yu Gu, Boyuan Zheng, Shijie Chen, Samuel Stevens, Boshi Wang, Huan Sun, and Yu Su. 2023. Mind2web: Towards a generalist agent for the web. Preprint, arXiv:2306.06070.
  6. Luyu Gao, Aman Madaan, Shuyan Zhou, Uri Alon, Pengfei Liu, Yiming Yang, Jamie Callan, and Graham Neubig. 2023. Pal: Program-aided language models. Preprint, arXiv:2211.10435. find
  7. Yifei Gao, Junhong Ye, Jiaqi Wang, and Jitao Sang. 2025. Websynthesis: World-model-guided mcts for efficient webui-trajectory synthesis. Preprint, arXiv:2507.04370. — [arXiv:2507.04370](https://arxiv.org/abs/2507.04370) find
  8. Yadong Lu, Wei-ge Chen, Olga Vrousgos, Corby Rosset, and 1 others. 2024. Agentinstruct: Toward generative teaching with agentic flows. arXiv preprint arXiv:2407.03502. find
  9. Tao Ge, Xin Chan, Xiaoyang Wang, Dian Yu, Haitao Mi, and Dong Yu. 2024. Scaling synthetic data creation with 1,000,000,000 personas. arXiv preprint arXiv:2406.20094.
  10. Shikhar Murty, Hao Zhu, Dzmitry Bahdanau, and Christopher D Manning. 2024. Nnetnav: Unsupervised learning of browser agents through environment interaction in the wild. arXiv preprint arXiv:2410.02907. find
  11. Daya Guo, Dejian Yang, Haowei Zhang, Junxiao Song, Ruoyu Zhang, Runxin Xu, Qihao Zhu, Shirong Ma, Peiyi Wang, Xiao Bi, and 1 others. 2025. Deepseek-r1: Incentivizing reasoning capability in llms via reinforcement learning. arXiv preprint arXiv:2501.12948.
  12. Hongliang He, Wenlin Yao, Kaixin Ma, Wenhao Yu, Hongming Zhang, Tianqing Fang, Zhenzhong Lan, and Dong Yu. 2024. Openwebvoyager: Building multimodal web agents via iterative real-world exploration, feedback and optimization. arXiv preprint arXiv:2410.19609.
  13. Sirui Hong, Mingchen Zhuge, Jonathan Chen, Xiawu Zheng, Yuheng Cheng, Ceyao Zhang, Jinlin Wang, Zili Wang, Steven Ka Shing Yau, Zijuan Lin, and 1 others. 2024. Metagpt: Meta programming for a multi-agent collaborative framework. International Conference on Learning Representations, ICLR.
  14. Kuan Li, Zhongwang Zhang, Huifeng Yin, Liwen Zhang, Litu Ou, Jialong Wu, Wenbiao Yin, Baixuan Li, Zhengwei Tao, Xinyu Wang, and 1 others. 2025. Websailor: Navigating super-human reasoning for web agent. arXiv preprint arXiv:2507.02592. find
  15. Hunter Lightman, Vineet Kosaraju, Yuri Burda, Harrison Edwards, Bowen Baker, Teddy Lee, Jan Leike, John Schulman, Ilya Sutskever, and Karl Cobbe. 2023. Let’s verify step by step. In The Twelfth International Conference on Learning Representations.
  16. Bang Liu, Xinfeng Li, Jiayi Zhang, Jinlin Wang, Tanjin He, Sirui Hong, Hongzhang Liu, Shaokun Zhang, Kaitao Song, Kunlun Zhu, and 1 others. 2025. Advances and challenges in foundation agents: From brain-inspired intelligence to evolutionary, collaborative, and safe systems. arXiv preprint arXiv:2504… — [arXiv:2504.01990](https://arxiv.org/abs/2504.01990) find
  17. Xing Han Lu, Zdeněk Kasner, and Siva Reddy. 2024. WebLINX: Real-world website navigation with multi-turn dialogue. In Forty-first International Conference on Machine Learning.
  18. Zhengxi Lu, Yuxiang Chai, Yaxuan Guo, Xi Yin, Liang Liu, Hao Wang, Han Xiao, Shuai Ren, Guanjing Xiong, and Hongsheng Li. 2025. Ui-r1: Enhancing efficient action prediction of gui agents by reinforcement learning. arXiv preprint arXiv:2503.21620. — [arXiv:2503.21620](https://arxiv.org/abs/2503.21620) find
  19. Arindam Mitra, Luciano Del Corro, Guoqing Zheng, Shweti Mahajan, Dany Rouhana, Andres Codas, Reiichiro Nakano, Jacob Hilton, Suchir Balaji, Jeff Wu, Long Ouyang, Christina Kim, Christopher Hesse, Shantanu Jain, Vineet Kosaraju, William Saunders, and 1 others. 2021. Webgpt: Browser-assisted question-…
  20. Tianyue Ou, Frank F Xu, Aman Madaan, Jiarui Liu, Robert Lo, Abishek Sridhar, Sudipta Sengupta, Dan Roth, Graham Neubig, and Shuyan Zhou. 2024. Synatra: Turning indirect knowledge into direct demonstrations for digital agents at scale. Advances in Neural Information Processing Systems, 37:91618– 9165… find
  21. Vardaan Pahuja, Yadong Lu, Corby Rosset, Boyu Gou, Arindam Mitra, Spencer Whitehead, Yu Su, and Ahmed Awadallah. 2025. Explorer: Scaling exploration-driven web trajectory synthesis for multimodal web agents. Preprint, arXiv:2502.11357. find
  22. Yujia Qin, Yining Ye, Junjie Fang, Haoming Wang, Shihao Liang, Shizuo Tian, Junda Zhang, Jiahao Li, Yunxin Li, Shijue Huang, and 1 others. 2025. Ui-tars: Pioneering automated gui interaction with native agents. arXiv preprint arXiv:2501.12326.
  23. Jiahao Qiu, Xuan Qi, Tongcheng Zhang, Xinzhe Juan, Jiacheng Guo, Yifu Lu, Yimin Wang, Zixin Yao, Qihan Ren, Xun Jiang, and 1 others. 2025. Alita: Generalist agent enabling scalable agentic reasoning with minimal predefinition and maximal self-evolution. arXiv preprint arXiv:2505.20286. — [arXiv:2505.20286](https://arxiv.org/abs/2505.20286) find
  24. Rafael Rafailov, Archit Sharma, Eric Mitchell, Christopher D Manning, Stefano Ermon, and Chelsea Finn. 2023. Direct preference optimization: Your language model is secretly a reward model. Advances in neural information processing systems, 36:53728–53741.
  25. Hongjin Su, Ruoxi Sun, Jinsung Yoon, Pengcheng Yin, Tao Yu, and Sercan Ö Arık. 2025. Learn-by-interact: A data-centric framework for self-adaptive agents in realistic environments. arXiv preprint arXiv:2501.10893. — [arXiv:2501.10893](https://arxiv.org/abs/2501.10893) find
  26. Qiushi Sun, Kanzhi Cheng, Zichen Ding, Chuanyang Jin, Yian Wang, Fangzhi Xu, Zhenyu Wu, Chengyou Jia, Liheng Chen, Zhoumianze Liu, Ben Kao, Guohao Li, Junxian He, Yu Qiao, and Zhiyong Wu. 2025. Os-genesis: Automating gui agent trajectory construction via reverse task synthesis. Preprint, arXiv:2412.…
  27. Rohan Taori, Ishaan Gulrajani, Tianyi Zhang, Yann Dubois, Xuechen Li, Carlos Guestrin, Percy Liang, and Tatsunori B. Hashimoto. 2023. Stanford alpaca: An instruction-following llama model. https://github.com/tatsu-lab/stanford_alpaca. ↗
  28. Laurens van der Maaten and Geoffrey Hinton. 2008. Visualizing data using t-sne. Journal of Machine Learning Research, 9(86):2579–2605. find
  29. Lei Wang, Chen Ma, Xueyang Feng, Zeyu Zhang, Hao Yang, Jingsen Zhang, Zhiyuan Chen, Jiakai Tang, Xu Chen, Yankai Lin, and 1 others. 2024. A survey on large language model based autonomous agents. Frontiers of Computer Science, 18(6):186345.
  30. Xinyuan Wang, Bowen Wang, Dunjie Lu, Junlin Yang, Tianbao Xie, Junli Wang, Jiaqi Deng, Xiaole Guo, Yiheng Xu, Chen Henry Wu, and 1 others. 2025. Opencua: Open foundations for computer-use agents. arXiv preprint arXiv:2508.09123. find
  31. Xuezhi Wang, Jason Wei, Dale Schuurmans, Quoc Le, Ed Chi, Sharan Narang, Aakanksha Chowdhery, and Denny Zhou. 2022. Self-consistency improves chain of thought reasoning in language models. arXiv preprint arXiv:2203.11171.
  32. Yizhong Wang, Yeganeh Kordi, Swaroop Mishra, Alisa Liu, Noah A. Smith, Daniel Khashabi, and Hannaneh Hajishirzi. 2023a. Self-instruct: Aligning language models with self-generated instructions. Preprint, arXiv:2212.10560. find
  33. Zhaoyang Wang, Shaohan Huang, Yuxuan Liu, Jiahai Wang, Minghui Song, Zihan Zhang, Haizhen Huang, Furu Wei, Weiwei Deng, Feng Sun, and 1 others. 2023b. Democratizing reasoning ability: Tailored learning from large language model. arXiv preprint arXiv:2310.13332. find
  34. Jason Wei, Zhiqing Sun, Spencer Papay, Scott McKinney, Jeffrey Han, Isa Fulford, Hyung Won Chung, Alex Tachard Passos, William Fedus, and Amelia Glaese. 2025a. Browsecomp: A simple yet challenging benchmark for browsing agents. arXiv preprint arXiv:2504.12516. find
  35. Jason Wei, Xuezhi Wang, Dale Schuurmans, Maarten Bosma, Fei Xia, Ed Chi, Quoc V Le, Denny Zhou, and 1 others. 2022. Chain-of-thought prompting elicits reasoning in large language models. Advances in neural information processing systems, 35:24824– 24837.
  36. Zhepei Wei, Wenlin Yao, Yao Liu, Weizhi Zhang, Qin Lu, Liang Qiu, Changlong Yu, Puyang Xu, Chao Zhang, Bing Yin, and 1 others. 2025b. Webagent-r1: Training web agents via end-to-end multi-turn reinforcement learning. arXiv preprint arXiv:2505.16421.
  37. Qingyun Wu, Gagan Bansal, Jieyu Zhang, Yiran Wu, Beibin Li, Erkang Zhu, Li Jiang, Xiaoyun Zhang, Shaokun Zhang, Jiale Liu, and 1 others. 2024. Autogen: Enabling next-gen llm applications via multi-agent conversations. In First Conference on Language Modeling. find
  38. Tianbao Xie, Danyang Zhang, Jixuan Chen, Xiaochuan Li, Siheng Zhao, Ruisheng Cao, Toh J Hua, Zhoujun Cheng, Dongchan Shin, Fangyu Lei, and 1 others. 2024. Osworld: Benchmarking multimodal agents for open-ended tasks in real computer environments. Advances in Neural Information Processing Systems, 37…
  39. Can Xu, Qingfeng Sun, Kai Zheng, Xiubo Geng, Pu Zhao, Jiazhan Feng, Chongyang Tao, and Daxin Jiang. 2023. Wizardlm: Empowering large language models to follow complex instructions. arXiv preprint arXiv:2304.12244.
  40. Yiheng Xu, Dunjie Lu, Zhennan Shen, Junli Wang, Zekun Wang, Yuchen Mao, Caiming Xiong, and Tao Yu. 2024. Agenttrek: Agent trajectory synthesis via guiding replay with web tutorials. arXiv preprint arXiv:2412.09605. — [arXiv:2412.09605](https://arxiv.org/abs/2412.09605) find
  41. Tianci Xue, Weijian Qi, Tianneng Shi, Chan Hee Song, Boyu Gou, Dawn Song, Huan Sun, and Yu Su. 2025. An illusion of progress? assessing the current state of web agents. In Second Conference on Language Modeling. find
  42. An Yang, Anfeng Li, Baosong Yang, Beichen Zhang, Binyuan Hui, Bo Zheng, Bowen Yu, Chang Gao, Chengen Huang, Chenxu Lv, and 1 others. 2025a. Qwen3 technical report. arXiv preprint arXiv:2505.09388.
  43. Ke Yang, Yao Liu, Sapana Chaudhary, Rasool Fakoor, Pratik Chaudhari, George Karypis, and Huzefa Rangwala. 2025b. Agentoccam: A simple yet strong baseline for LLM-based web agents. In The Thirteenth International Conference on Learning Representations. find
  44. Shunyu Yao, Howard Chen, John Yang, and Karthik Narasimhan. 2022. Webshop: Towards scalable real-world web interaction with grounded language agents. Advances in Neural Information Processing Systems, 35:20744–20757.
  45. Shunyu Yao, Dian Yu, Jeffrey Zhao, Izhak Shafran, Tom Griffiths, Yuan Cao, and Karthik Narasimhan. 2023a. Tree of thoughts: Deliberate problem solving with large language models. Advances in neural information processing systems, 36:11809–11822. find
  46. Shunyu Yao, Jeffrey Zhao, Dian Yu, Nan Du, Izhak Shafran, Karthik Narasimhan, and Yuan Cao. 2023b. React: Synergizing reasoning and acting in language models. Preprint, arXiv:2210.03629.
  47. Eric Zelikman, Yuhuai Wu, Jesse Mu, and Noah Goodman. 2022. Star: Bootstrapping reasoning with reasoning. Advances in Neural Information Processing Systems, 35:15476–15488.
  48. Ruichen Zhang, Mufan Qiu, Zhen Tan, Mohan Zhang, Vincent Lu, Jie Peng, Kaidi Xu, Leandro Z Agudelo, Peter Qian, and Tianlong Chen. 2025. Symbiotic cooperation for web agents: Harnessing complementary strengths of large and small llms. arXiv preprint arXiv:2502.07942. — [arXiv:2502.07942](https://arxiv.org/abs/2502.07942) find
  49. Andrew Zhao, Yiran Wu, Yang Yue, Tong Wu, Quentin Xu, Matthieu Lin, Shenzhi Wang, Qingyun Wu, Zilong Zheng, and Gao Huang. 2025. Absolute zero: Reinforced self-play reasoning with zero data. arXiv preprint arXiv:2505.03335. find
  50. Lianmin Zheng, Wei-Lin Chiang, Ying Sheng, Siyuan Zhuang, Zhanghao Wu, Yonghao Zhuang, Zi Lin, Zhuohan Li, Dacheng Li, Eric Xing, Hao Zhang, Joseph E Gonzalez, and Ion Stoica. 2023. Judging llm-as-a-judge with mt-bench and chatbot arena. In Advances in Neural Information Processing Systems, volume 3…
  51. Shuyan Zhou, Frank F. Xu, Hao Zhu, Xuhui Zhou, Robert Lo, Abishek Sridhar, Xianyi Cheng, Tianyue Ou, Yonatan Bisk, Daniel Fried, Uri Alon, and Graham Neubig. 2024. Webarena: A realistic web environment for building autonomous agents. Preprint, arXiv:2307.13854.

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/wang-2026-synthagent-adapting-web-agents-synthetic/