TY  - RPRT
TI  - MSPR: Multi-scale Predictive Representations for Goal-conditioned Reinforcement Learning
AU  - Valliappan Chidambaram Adaikkappan
AU  - Sai Rajeswar
AU  - Pietro Mazzaglia
AU  - David Meger
PY  - 2026
UR  - https://arxiv.org/abs/2605.09364
ID  - 2605.09364
ER  -