TY  - RPRT
TI  - Learning How Hard to Think: Input-Adaptive Allocation of LM Computation
AU  - Mehul Damani
AU  - Idan Shenfeld
AU  - Andi Peng
AU  - Andreea Bobu
AU  - Jacob Andreas
PY  - 2024
UR  - https://arxiv.org/abs/2410.04707
ID  - 2410.04707
ER  -