OCR-Memory: Optical Context Retrieval for Long-Horizon Agent Memory
Jinze Li, Yang Zhang, Xin Yang, Jiayi Qu, Jinfeng Xu, Shuo Yang, Junhua Ding, Edith Cheuk-Han Ngai · 2026
Read OCR-Memory: Optical Context Retrieval for Long-Horizon Agent Memory by Jinze Li, Yang Zhang, Xin Yang, Jiayi Qu, Jinfeng Xu, Shuo Yang, Junhua Ding, Edith Cheuk-Han Ngai (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-26622.txt.
Description
Li and seven colleagues propose OCR-Memory, a memory framework that stores agent experience as images rather than raw text. Long-horizon agents need past experience, but storing full trajectories costs too many tokens, while summaries and text retrieval lose information and fragment evidence. OCR-Memory renders historical trajectories into images marked with unique visual identifiers, then retrieves them by locating visual anchors and transcribing the matching verbatim text. Experiments on long-horizon benchmarks show steady gains under strict context limits.
Catalog record
| CREATOR | Jinze Li, Yang Zhang, Xin Yang, Jiayi Qu, Jinfeng Xu, Shuo Yang, Junhua Ding, Edith Cheuk-Han Ngai |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 41294 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.26622