TY  - RPRT
TI  - The Imperfective Paradox Is Not Necessarily in Large Language Models: A Benchmark Failure Before a Model Failure
AU  - Kaiqiao Han
AU  - Yizhou Sun
PY  - 2026
UR  - https://arxiv.org/abs/2608.25005
ID  - 2608.25005
ER  -