IAAILoopy
Loopy: Low-Bit Quantization Framework for Looped Language Models
A post-training quantization framework that selects shared-core low-bit representations by their final prediction loss at the target deployment depth, because the ranking of quantization configurations changes with recurrent depth. Calibration windows are allocated progressively to promising candidates using forward passes only.
Inside the method
Candidate rotations / scalings → evaluate at target depth R → shared low-bit core
Simplified conceptual schematic. Consult the paper for the complete architecture.
- Recurrence family
- Full-stack recurrence
- Depth control
- Quantization tuned for a target deployment depth
- KV / state strategy
- See paper
Reading note
State of the art among compared baselines in eight settings; on Ouro-1.4B under W4A4 it lowers LAMBADA perplexity by 36.5% relative to SpinQuant.
Sources checked 2026-10-28. This catalog does not imply independent reproduction.
Related works
Cite this work
@misc{loopy2026,
title = {Loopy: Low-Bit Quantization Framework for Looped Language Models},
author = {Zeyu Li and Yipu Zhang and Jintao Cheng and Xin Li and Wei Zhnag},
year = {2026},
eprint = {2610.05265},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.05265}
}