Quantization Robustness of Monotone Operator Equilibrium Networks

James Li, Philip H.W. Leong, Thomas Chaffey · 2026

Read Quantization Robustness of Monotone Operator Equilibrium Networks by James Li, Philip H.W. Leong, Thomas Chaffey (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-2603-10562.txt.

Description

James Li, Philip H W Leong, and Thomas Chaffey study what happens when an implicit layer model runs on low precision hardware. These models output the single fixed point of a monotone operator, which guarantees a solution exists and can be found. Quantized weights can break those guarantees. Treating quantization as a small spectral change, they show the solver still converges when the weight change stays below the monotonicity margin. They bound how far the fixed point moves, and MNIST tests showed the predicted phase change across bit widths.

Catalog record

CREATORJames Li, Philip H.W. Leong, Thomas Chaffey
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE34834
LICENSEhttp://creativecommons.org/licenses/by-sa/4.0/
RIGHTSCC BY SA
SOURCEarXiv

Source: https://arxiv.org/abs/2603.10562

Browse the full library · About the Tower of Babel