# Mode-Dependent Rectification for Stable PPO Training

arXiv: https://arxiv.org/abs/2602.05619 · HTML: https://mlsanity.com/paper/2602.05619/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Placed in RL theory and sample complexity, one of 15 directions in Deep reinforcement learning methods.
- Builds on 26 of its 36 references that are in the map.
- First posted on arXiv 5 Feb 2026.

## Placed in

- Deep reinforcement learning methods / [RL theory and sample complexity](https://mlsanity.com/deeprl/rl-theory-and-sample-complexity/)

## Builds on (most cited first, 26 of 26)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [Adam: A Method for Stochastic Optimization](https://mlsanity.com/paper/1412.6980/) (2014)
- [Proximal Policy Optimization Algorithms](https://mlsanity.com/paper/1707.06347/) (2017)
- [Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift](https://mlsanity.com/paper/1502.03167/) (2015)
- [Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor](https://mlsanity.com/paper/1801.01290/) (2018)
- [Continuous control with deep reinforcement learning](https://mlsanity.com/paper/1509.02971/) (2015)
- [Playing Atari with Deep Reinforcement Learning](https://mlsanity.com/paper/1312.5602/) (2013)
- [Addressing Function Approximation Error in Actor-Critic Methods](https://mlsanity.com/paper/1802.09477/) (2018)
- [High-Dimensional Continuous Control Using Generalized Advantage Estimation](https://mlsanity.com/paper/1506.02438/) (2015)
- [Group Normalization](https://mlsanity.com/paper/1803.08494/) (2018)
- [Instance Normalization: The Missing Ingredient for Fast Stylization](https://mlsanity.com/paper/1607.08022/) (2016)
- [IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures](https://mlsanity.com/paper/1802.01561/) (2018)
- [Leveraging Procedural Generation to Benchmark Reinforcement Learning](https://mlsanity.com/paper/1912.01588/) (2019)
- [Quantifying Generalization in Reinforcement Learning](https://mlsanity.com/paper/1812.02341/) (2018)
- [Randomized Ensembled Double Q-Learning: Learning Fast Without a Model](https://mlsanity.com/paper/2101.05982/) (2021)
- [Revisiting the Arcade Learning Environment: Evaluation Protocols and Open Problems for General Agents](https://mlsanity.com/paper/1709.06009/) (2017)
- [TinyViT: Fast Pretraining Distillation for Small Vision Transformers](https://mlsanity.com/paper/2207.10666/) (2022)
- [Batch Renormalization: Towards Reducing Minibatch Dependence in Batch-Normalized Models](https://mlsanity.com/paper/1702.03275/) (2017)
- [Dropout Q-Functions for Doubly Efficient Reinforcement Learning](https://mlsanity.com/paper/2110.02034/) (2021)
- [CrossQ: Batch Normalization in Deep Reinforcement Learning for Greater Sample Efficiency and Simplicity](https://mlsanity.com/paper/1902.05605/) (2019)
- [Normalization and effective learning rates in reinforcement learning](https://mlsanity.com/paper/2407.01800/) (2024)
- [CrossNorm and SelfNorm for Generalization under Distribution Shifts](https://mlsanity.com/paper/2102.02811/) (2021)
- [A Study of Plasticity Loss in On-Policy Deep Reinforcement Learning](https://mlsanity.com/paper/2405.19153/) (2024)
- [Normalization Enhances Generalization in Visual Reinforcement Learning](https://mlsanity.com/paper/2306.00656/) (2023)
- [Rogue-Gym: A New Challenge for Generalization in Reinforcement Learning](https://mlsanity.com/paper/1904.08129/) (2019)

## Built on by (most cited first, 1 of 1)

- [Interactive Whole Slide Images for RL-based Tumour Segmentation](https://mlsanity.com/paper/2608.16607/) (2026)

## Abstract (arXiv)

Mode-dependent architectural components (layers that behave differently during training and evaluation, such as Batch Normalization or dropout) are commonly used in visual reinforcement learning but can destabilize on-policy optimization. We show that in Proximal Policy Optimization (PPO), discrepancies between training and evaluation behavior induced by Batch Normalization lead to policy mismatch, distributional drift, and reward collapse. We propose Mode-Dependent Rectification (MDR), a lightweight dual-phase training procedure that stabilizes PPO under mode-dependent layers without architectural changes. Experiments across procedurally generated games and real-world patch-localization tasks demonstrate that MDR consistently improves stability and performance, and extends naturally to other mode-dependent layers.