@inproceedings{feng-etal-2026-agentocr,
    title = "{A}gent{OCR}: Reimagining Agent History via Optical Self-Compression",
    author = "Feng, Lang  and
      Yang, Fuchao  and
      Chen, Feng  and
      Cheng, Xin  and
      Xu, Haiyang  and
      Wan, Zhenglin  and
      Yan, Ming  and
      An, Bo",
    editor = "Liakata, Maria  and
      Moreira, Viviane P.  and
      Zhang, Jiajun  and
      Jurgens, David",
    booktitle = "Proceedings of the 64th Annual Meeting of the {A}ssociation for {C}omputational {L}inguistics (Volume 1: Long Papers)",
    month = jul,
    year = "2026",
    address = "San Diego, California, United States",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingest-acl/2026.acl-long.230/",
    pages = "5067--5086",
    ISBN = "979-8-89176-390-6",
    abstract = "Recent advances in large language models (LLMs) enable agentic systems trained with reinforcement learning (RL) over multi-turn interaction, but practical deployment is bottlenecked by rapidly growing textual histories that inflate token and memory costs. We introduce AgentOCR, a framework that exploits visual tokens' superior information density by representing the accumulated observation-action history as a compact rendered image. To make multi-turn rollouts scalable, AgentOCR proposes segment optical caching. By decomposing history into hashable segments and maintaining a visual cache, this mechanism eliminates redundant re-rendering. Beyond fixed rendering, AgentOCR introduces agentic self-compression, where the agent actively emits a compression rate and is trained with compression-aware reward to adaptively balance task success and token efficiency. We conduct extensive experiments on challenging agentic benchmarks, ALFWorld and search-based QA. Remarkably, AgentOCR preserves over 95{\%} of text-based agent performance while substantially reducing token consumption ({\ensuremath{>}}50{\%}), yielding consistent token and memory efficiency. Further analysis validates a 20$\times$ rendering speedup from optical caching and effective self-compression balancing. Our code is available at https://github.com/langfengQ/AgentOCR."
}