TY  - RPRT
TI  - Preference learning in shades of gray: Interpretable and bias-aware reward modeling for human preferences
AU  - Simona-Vasilica Oprea
AU  - Adela Bâra
PY  - 2026
UR  - https://arxiv.org/abs/2604.01312
ID  - 2604.01312
ER  -