IncreFA: Breaking the Static Wall of Generative Model Attribution
Haotian Qin, Dongliang Chang, Yueying Gao, Yuexuan Tan, Lei Chen, Zhanyu Ma · 2026
Read IncreFA: Breaking the Static Wall of Generative Model Attribution by Haotian Qin, Dongliang Chang, Yueying Gao, Yuexuan Tan, Lei Chen, Zhanyu Ma (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-2604-17736.txt.
Description
Qin and five colleagues introduce IncreFA, framing credit as ordered stepwise learning, so the system keeps learning as new models emerge. Image credit is a moving target as new generators appear monthly, making watermark, sorter, and reverse methods dead. The core problem is not spotting but the failure to adapt credit itself. IncreFA uses layered links among generative designs. It joins two parts. Hierarchical Constraints encode design layers through tuned split priors, splitting family-level constants from model-specific quirks. A Latent Memory Bank replays compact codes.
Catalog record
| CREATOR | Haotian Qin, Dongliang Chang, Yueying Gao, Yuexuan Tan, Lei Chen, Zhanyu Ma |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 39624 |
| LICENSE | http://creativecommons.org/publicdomain/zero/1.0/ |
| RIGHTS | CC0 |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.17736