A simple PPO training scheme that adds hidden layers over time via function-preserving Net2Net morphisms outperforms static networks of the same final depth on MiniHack Room and MuJoCo Ant.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Growing with Experience: Growing Neural Networks in Deep Reinforcement Learning
A simple PPO training scheme that adds hidden layers over time via function-preserving Net2Net morphisms outperforms static networks of the same final depth on MiniHack Room and MuJoCo Ant.