TY  - RPRT
TI  - Unveiling Reasoning Thresholds in Language Models: Scaling, Fine-Tuning, and Interpretability through Attention Maps
AU  - Yen-Che Hsiao
AU  - Abhishek Dutta
PY  - 2025
UR  - https://arxiv.org/abs/2502.15120
ID  - 2502.15120
ER  -