TY  - RPRT
TI  - Human-Inspired Learning for Large Language Models via Obvious Record and Maximum-Entropy Method Discovery
AU  - Hong Su
PY  - 2025
UR  - https://arxiv.org/abs/2512.12608
ID  - 2512.12608
ER  -