fix citations

Author: kameronton

paper.bib

@@ -66,12 +66,11 @@ @article{raffin2021stable
   url = {http://jmlr.org/papers/v22/20-1364.html}
 }
 
-@article{towers2026gymnasium,
-  title={Gymnasium: A standard interface for reinforcement learning environments},
-  author={Towers, Mark and Kwiatkowski, Ariel and Balis, John and De Cola, Gianluca and Deleu, Tristan and Goul{\~a}o, Manuel and Andreas, Kallinteris and Krimmel, Markus and Kg, Arjun and Perez-Vicente, Rodrigo and others},
-  journal={Advances in Neural Information Processing Systems},
-  volume={38},
-  year={2026}
+@article{towers2024gymnasium,
+  title={Gymnasium: A Standard Interface for Reinforcement Learning Environments},
+  author={Towers, Mark and Kwiatkowski, Ariel and Terry, Jordan and Balis, John U and De Cola, Gianluca and Deleu, Tristan and Goul{\~a}o, Manuel and Kallinteris, Andreas and Krimmel, Markus and KG, Arjun and others},
+  journal={arXiv preprint arXiv:2407.17032},
+  year={2024}
 }
 
 @inproceedings{verma2018programmatically,
@@ -82,16 +81,23 @@ @inproceedings{verma2018programmatically
   year = {2018},
   organization = {PMLR}
 }
-
-@article{silva2020optimization,
-  title = {Optimization Methods for Interpretable Differentiable Decision Trees Applied to Reinforcement Learning},
-  author = {Andrew Silva and Matthew Gombolay and Taylor Killian and Ivan Jimenez and Sung-Hyun Son},
-  journal = {Proceedings of the International Conference on Artificial Intelligence and Statistics},
-  pages = {1855--1865},
-  year = {2020},
-  organization = {PMLR}
+@InProceedings{silva2020optimization,
+  title = 	 {Optimization Methods for Interpretable Differentiable Decision Trees Applied to Reinforcement Learning},
+  author =       {Silva, Andrew and Killian, Taylor and Jimenez, Ivan and Son, Sung-Hyun and Gombolay, Matthew},
+  booktitle = 	 {Proceedings of the Twenty Third International Conference on Artificial Intelligence and Statistics},
+  pages = 	 {1855--1865},
+  year = 	 {2020},
+  editor = 	 {Chiappa, Silvia and Calandra, Roberto},
+  volume = 	 {108},
+  series = 	 {Proceedings of Machine Learning Research},
+  month = 	 {26--28 Aug},
+  publisher =    {PMLR},
+  pdf = 	 {http://proceedings.mlr.press/v108/silva20a/silva20a.pdf},
+  url = 	 {https://proceedings.mlr.press/v108/silva20a.html},
+  abstract = 	 { Decision trees are ubiquitous in machine learning for their ease of use and interpretability. Yet, these models are not typically employed in reinforcement learning as they cannot be updated online via stochastic gradient descent. We overcome this limitation by allowing for a gradient update over the entire tree that improves sample complexity affords interpretable policy extraction. First, we include theoretical motivation on the need for policy-gradient learning by examining the properties of gradient descent over differentiable decision trees. Second, we demonstrate that our approach equals or outperforms a neural network on all domains and can learn discrete decision trees online with average rewards up to 7x higher than a batch-trained decision tree. Third, we conduct a user study to quantify the interpretability of a decision tree, rule list, and a neural network with statistically significant results (p < 0.001).}
 }
 
+
 @inproceedings{panda2024vanilla,
   author       = {Subrat Prasad Panda and
                   Blaise Genest and

paper.md

@@ -26,7 +26,7 @@ bibliography: paper.bib
 BordAX is a Python framework for reinforcement learning (RL) built on JAX [@jax2018github].
 It provides a modular, fully JIT-compilable training pipeline that supports multiple policy representations, including standard multilayer perceptrons (MLPs), differentiable decision trees (DTSemNet, [@panda2024vanilla]), and boolean function networks (HyperBool).
 BordAX currently implements Proximal Policy Optimization (PPO) [@schulman2017proximal] and Deep Q-Networks (DQN) [@mnih2015human], and is designed so that additional algorithms can be composed from interchangeable collector, batch-builder, and updater components.
-The framework supports both Gymnax [@gymnax2022github] environments, which enable full JIT compilation of the training loop via `jax.lax.scan`, and standard Gymnasium [@towers2026gymnasium] environments.
+The framework supports both Gymnax [@gymnax2022github] environments, which enable full JIT compilation of the training loop via `jax.lax.scan`, and standard Gymnasium [@towers2024gymnasium] environments.
 On a CartPole-v1 benchmark with identical hyperparameters, BordAX with Gymnax achieves approximately 2.8 times higher throughput than Stable-Baselines3 [@raffin2021stable].
 
 # Statement of need