Reproduce “Instance-Fitted Representations for Data Compression: A Survey” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Youneng Bao, Yulong Cheng, Jiacong Chen, Yiping Liu, Qingyu Mao, Mu Li, Chuanmin Jia, Li Song and 2 more
The survey defines instance-fitted representation compression (IFRC) as fitting a representation to one target and reconstructing from its fitted state. It distinguishes functional representations from parametric primitives and separates model proxies, estimates, and complete serialized payloads. Its framework says only complete payloads under aligned conditions support cross-paper rate–distortion conclusions; its examples do not establish a universal winner.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.