diff --git a/README.md b/README.md
index 0634a85e..2a69b254 100644
--- a/README.md
+++ b/README.md
@@ -85,6 +85,7 @@ pip install -r requirements/requirements-pettingzoo.txt
 pip install -r requirements/requirements-jax.txt
 pip install -r requirements/requirements-docs.txt
 pip install -r requirements/requirements-cloud.txt
+pip install -r requirements/requirements-memory_gym.txt
 ```
 
 To run training scripts in other games:
diff --git a/docs/rl-algorithms/ppo-trxl.md b/docs/rl-algorithms/ppo-trxl.md
index 55a49d53..dee80a55 100644
--- a/docs/rl-algorithms/ppo-trxl.md
+++ b/docs/rl-algorithms/ppo-trxl.md
@@ -7,7 +7,9 @@ Real-world tasks may expose imperfect information (e.g. partial observability).
 Original Paper and Implementation
 
 * :material-file-document: [Memory Gym: Towards Endless Tasks to Benchmark Memory Capabilities of Agents](https://arxiv.org/abs/2309.17207)
-* :material-github: [neroRL](https://github.com/MarcoMeter/neroRL), [Episodic Transformer Memory PPO](https://github.com/MarcoMeter/episodic-transformer-memory-ppo)
+* :material-github: [neroRL](https://github.com/MarcoMeter/neroRL)
+* :material-github:  [Episodic Transformer Memory PPO](https://github.com/MarcoMeter/episodic-transformer-memory-ppo)
+* :material-github: [Endless Memory Gym](https://github.com/MarcoMeter/endless-memory-gym)
 * :material-play-circle: [Interactive Visualizations of Trained Agents](https://marcometer.github.io/)
 
 Related Publications and Repositories