tensorflow/lucid

Health v2 · maintenance only

Flags: archived

Adoption not part of the score

4704 stars · 642 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

Lucid is a collection of infrastructure and tools for research in neural network interpretability, built on TensorFlow 1.x. It provides feature visualization techniques (like DeepDream-style optimization) and runnable Jupyter/Colab notebooks for visualizing and understanding neural networks.

Use cases

• visualize what neurons in a neural network have learned
• run feature visualization experiments like DeepDream
• interpret and explain convolutional neural network activations
• explore neural network interpretability in Colab without setup
• visualize my own TensorFlow model's internal features
• learn feature visualization techniques through tutorial notebooks

When to choose

• you research neural network interpretability or feature visualization
• you work with TensorFlow 1.x models and want to visualize their internals
• you want ready-to-run Colab notebooks for interpretability experiments

When to avoid

• you need TensorFlow 2.x support, which Lucid does not currently support
• you need production-grade, well-supported software with technical support
• you need interpretability tools for PyTorch or other frameworks

Facets

library · maturity maintenance

machine-learning data-visualization deep-learning machine-learning deep-learning data-visualization artificial-intelligence python browser cross-platform feature-visualization interpretability tensorflow deepdream neural-networks colab research-code

Member repositories

For agents

markdown · JSON · MCP: product_card(name="tensorflow/lucid")