Show Your Work
==============

We'd love to see what you've built with SymTorch!

Why Use SymTorch?
-----------------

SymTorch bridges the gap between the power of deep learning and the interpretability of symbolic mathematics. By combining neural networks with symbolic regression, you can:

- **Discover interpretable equations** from your trained models
- **Understand what your neural networks have learned** at both layer and model levels
- **Extract symbolic knowledge** from black-box models
- **Create explainable AI systems** that are both accurate and interpretable
- **Accelerate scientific discovery** by uncovering mathematical relationships in your data

Whether you're working on physics-informed neural networks, graph neural networks, or classical machine learning problems, SymTorch provides the tools to transform learned representations into human-readable mathematical expressions.

Featured Projects
-----------------

.. grid:: 3

   .. grid-item-card:: Symbolic Surrogates of Transformer MLPs
      :link: https://github.com/elizabethsztan/LLM_PCA

      .. image:: _static/symtorch_gallery/llm_pca.png

      +++
      Novel proof-of-concept framework to increase LLM throughput by replacing MLPs with symbolic expressions.

   .. grid-item-card:: Recovering Physical Laws from GNNs
      :link: https://github.com/elizabethsztan/SymTorch_symbolic_distillation_GNNs

      .. image:: _static/symtorch_gallery/gnn.png

      +++
      Detailed reproduction of the `paper <https://arxiv.org/abs/2006.11287>`_ with SymTorch.

   .. grid-item-card:: Upcoming Work
      :link: https://github.com/elizabethsztan/SymTorch

      .. image:: _static/symtorch_gallery/more_work.png

      +++
      Your work could go here!

Share Your Work With Us
-----------------------

Have you used SymTorch in your research, project, or application? We want to hear from you!

**We're looking for:**

- Research papers using SymTorch
- Industry applications and case studies
- Educational materials and tutorials
- Novel use cases and creative applications
- Performance benchmarks and comparisons
- Integration with other tools and frameworks

**What we'll showcase:**

- Links to your papers, repositories, or blog posts
- Brief descriptions of your application
- Key results and insights
- Visualizations or demonstrations (if available)
- Your contact information (if you'd like to share it)

How to Get Featured
-------------------

If you've used SymTorch and would like to share your work:

1. **Contact us** with details about your project
2. **Provide** a brief description (2-3 paragraphs) of your work
3. **Include** links to papers, code repositories, or demos
4. **Share** any visualizations or results (optional but appreciated)

**Contact Information:**

- **GitHub Issues**: `Open an issue <https://github.com/SymTorch/SymTorch/issues>`_ with the label "show-your-work"
- **Email**: Contact the maintainers directly through GitHub
- **Pull Request**: Add your project directly to this page via PR

Benefits of Sharing
-------------------

By sharing your work, you'll:

- **Gain visibility** for your research or project within the community
- **Inspire others** to explore new applications of symbolic regression
- **Contribute** to the growing ecosystem of interpretable machine learning
- **Connect** with other researchers and practitioners in the field
- **Help shape** the future development of SymTorch based on real-world use cases

Example Use Cases
-----------------

SymTorch is versatile and can be applied to various domains:

**Scientific Computing**
  - Discovering physical laws from experimental data
  - Simplifying complex physics-informed neural networks (PINNs)
  - Extracting governing equations from dynamical systems

**Machine Learning**
  - Creating interpretable alternatives to black-box models
  - Understanding intermediate representations in deep networks
  - Model compression through symbolic approximation

**Domain-Specific Applications**
  - Climate modeling and weather prediction
  - Financial modeling and risk analysis
  - Biological systems and computational biology
  - Materials science and chemistry

We Want to Hear Your Story
---------------------------

Every application of SymTorch is unique, and we're excited to learn how you're using symbolic regression to make neural networks more interpretable.

Whether you're a researcher publishing a paper, a student working on a class project, or an industry practitioner solving real-world problems, your work matters to the community.

**Don't hesitate to reach out** - we'd love to feature your work here!

---

*Last updated: 2025*
