LLM quantization

FIMA-Q

FIMA-Q: Post-Training Quantization for Vision Transformers by Fisher Information Matrix Approximation

Superseded baseline#28 of 80 most-superseded · first seen Jun 13, 2025

Superseded — cited as a baseline and beaten by newer methods

2 papers critique it · 1 beat it on benchmarks

What papers say

Verbatim critique sentences, each from a paper that cites FIMA-Q as a baseline.

FIMA-Q wu2025fima proposes an efficient Fisher Information Matrix approximation to guide quantization via block-wise reconstruction loss, pushing ViT PTQ toward W3A3 and W4A4 with notable improvement. However, the block-level independence neglects inter-block compensation, leaving further potential unexploited.
Joint Post-Training Quantization of Vision Transformers with Learned Prompt-Guided Data Generation
These strategies are largely layer or block-centric and optimize average fidelity. In COD at W4A4, the dominant failure is token-local: token-wise activation heterogeneity lets background tokens dominate the range and increase the zero-bin mass, which layer-wise fitting or reconstruction does not explicitly bound.
When W4A4 Breaks Camouflaged Object Detection: Token-Group Dual-Constraint Activation Quantization

Beaten on benchmarks

Head-to-head results where a newer method reports beating FIMA-Q. Values are copied from the source paper's tables — verify against the cited paper.

What to use instead

Recent methods in the same sub-problem, not yet superseded in the knowledge base — arXiv benchmark leaders, not vetted production recommendations.