One Token Away from Collapse: The Fragility of Instruction-Tuned Helpfulness

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Potraghloo, Erfan Baghaei, Azizi, Seyedarmin, Kundu, Souvik, Pedram, Massoud
Format:	Preprint
Published:	2026
Subjects:	Computation and Language Artificial Intelligence
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Internet

https://arxiv.org/abs/2604.13006