TY  - RPRT
TI  - Pruning via Causal Attribution Preserves Reasoning Performance in Large Language Models
AU  - Amogh Sheth
AU  - Biruk Assefa
AU  - Yi Wen Huang
AU  - Andrew Lin
AU  - Yuhao Ge
PY  - 2026
UR  - https://arxiv.org/abs/2606.19350
ID  - 2606.19350
ER  -