TY  - RPRT
TI  - A Convergence Framework for Deep $V$-Learning: Error Propagation and Sharp Action-Gap Bounds
AU  - Yury Kolomeytsev
PY  - 2026
UR  - https://arxiv.org/abs/2609.18782
ID  - 2609.18782
ER  -