TY  - RPRT
TI  - MIRA: Mitigating Pole Geometry Bias in Recurrent Language Models
AU  - Róisín Luo
AU  - Ihsan Ullah
AU  - Christian Gagné
AU  - Dong Li
AU  - Karyn Morrissey
PY  - 2026
UR  - https://arxiv.org/abs/2611.09412
ID  - 2611.09412
ER  -