Potato 2.0 at ACL 2026

The Potato 2.0 paper is published in the ACL 2026 System Demonstrations, and you can read it on the ACL Anthology. The paper describes Potato 2.0, with 39 annotation task types across text, audio, image, and video, support for labeling the outputs of AI agents, and an AI-in-the-loop workflow in which one annotator works alongside an LLM through iterative prompt refinement. The current release, Potato 2.9.1, supports 61 annotation types. If you use Potato in your research, cite the ACL 2026 paper.

Potato 2.0 additions beyond text annotation

The first Potato paper, at EMNLP 2022, described a config-first tool for text annotation. You write a YAML file and get a working interface without custom front-end code. The idea still holds in Potato 2.0, but the tool has grown well past text since then. The paper reports on three additions that came with that growth, and the feature docs cover each in more detail:

• Task types across modalities. The paper's 39 task types cover text, audio, image, and video, so one config can describe a sentiment scale, an audio event timeline, a bounding-box task, or a video rating.
• Labeling agent outputs. As people started evaluating LLMs and agents, Potato added displays for agent trajectories, tool calls, and multi-step outputs that annotators can label directly.
• AI-in-the-loop annotation. A human annotator and an LLM can work the same task together, with the human refining prompts as they go rather than labeling every item by hand.

Potato's publication record

Potato has appeared in the research record a few times, and for most uses you should cite the ACL 2026 paper by Jurgens, Chen, and Iyer, published in Volume 3 (System Demonstrations). The original paper by Pei et al. appeared in the EMNLP 2022 System Demonstrations and is also on arXiv. The Potato demo also received the Best Demo Award at HCOMP 2024.

BibTeX for citing Potato

Use this BibTeX entry to cite the ACL 2026 paper:

@inproceedings{jurgens-etal-2026-potato,
    title = "Potato 2.0: A Comprehensive Annotation Platform with {AI}-in-the-Loop Support",
    author = "Jurgens, David and Chen, Michael and Iyer, Lina",
    booktitle = "Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)",
    month = jul,
    year = "2026",
    address = "San Diego, California, United States",
    publisher = "Association for Computational Linguistics",
    pages = "374--386",
    url = "https://aclanthology.org/2026.acl-demo.37/",
}

If you are citing the original text-annotation work specifically, the EMNLP 2022 paper is still the right reference:

@inproceedings{pei-etal-2022-potato,
    title = "{POTATO}: The Portable Text Annotation Tool",
    author = "Pei, Jiaxin and Ananthasubramaniam, Aparna and Wang, Xingyao and Zhou, Naitian and Dedeloudis, Apostolos and Sargent, Jackson and Jurgens, David",
    booktitle = "Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing: System Demonstrations",
    month = dec,
    year = "2022",
    address = "Abu Dhabi, UAE",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2022.emnlp-demos.33/",
    pages = "327--337",
}

Thanks to everyone who has used Potato, filed issues, and sent annotation tasks we had not thought of. If you build something with Potato, we would like to hear about it on GitHub.

References

David Jurgens, Michael Chen, and Lina Iyer (2026). Potato 2.0: A Comprehensive Annotation Platform with AI-in-the-Loop Support. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations). https://aclanthology.org/2026.acl-demo.37/

Jiaxin Pei et al. (2022). POTATO: The Portable Text Annotation Tool. Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing: System Demonstrations. https://aclanthology.org/2022.emnlp-demos.33/