TY  - RPRT
TI  - Language Models as Zero-shot Lossless Gradient Compressors: Towards General Neural Parameter Prior Models
AU  - Hui-Po Wang
AU  - Mario Fritz
PY  - 2025
UR  - https://arxiv.org/abs/2409.17836
ID  - 2409.17836
ER  -