TY  - RPRT
TI  - Test-Time Scaling in Reasoning Models Is Not Effective for Knowledge-Intensive Tasks Yet
AU  - James Xu Zhao
AU  - Bryan Hooi
AU  - See-Kiong Ng
PY  - 2026
UR  - https://arxiv.org/abs/2509.06861
ID  - 2509.06861
ER  -