TY  - RPRT
TI  - The Maskability Index: Predicting Task-Objective Alignment in Pretrained Language Models
AU  - Ahmad Pouramini
AU  - Mahsa Afsharzadeh
PY  - 2026
UR  - https://arxiv.org/abs/2607.20265
ID  - 2607.20265
ER  -