Parameter-efficient fine-tuning (LoRA family)

raw gradient norm

Superseded baseline#903 of 1,113 most-superseded

Cited as a baseline — critiqued by newer work, not yet beaten on a benchmark here

1 papers critique it · 0 beat it on benchmarks

What papers say

Verbatim critique sentences, each from a paper that cites raw gradient norm as a baseline.

Raw gradient norm has been proposed as a layer-selection signal liu2025efficient, but can be noisy zhang23
Understanding and Guiding Layer Placement in Parameter-Efficient Fine-Tuning of Large Language Models

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Recent methods in the same sub-problem, not yet superseded in the knowledge base — arXiv benchmark leaders, not vetted production recommendations.