Christopher Olah

For most purposes, we discourage the use of PDFs because they will no longer be interactive. If you really need one, use your browsers print -> save as pdf option....

Hi Yang! [A Learned Representation For Artistic Style](https://arxiv.org/pdf/1610.07629.pdf) was published shortly afterwards and finds that it has a significant impact for feedforward style transfer. [Conditional Image Synthesis with Auxiliary Classifier...

What error do you get when you call render?

```Attempted to map inputs that were not found in graph_def: [input:0]``` When you exported your model, you had to declare what the name of the input node to your model...

When you run feature visualization, you optimize an image of size set by the argument `param_f`. Some models are incompatible with certain input sizes. Try doing something like: ``` render.render_vis(model,...

@ricardobarroslourenco @tschwabe @ebwendol @MahrRah @hegman12 We've been working on a new workflow for importing models into Lucid, which aims to make it much easier! We'd love your feedback. You can...

## Hypotheses: Why do poly-semantic neurons occur? ### The Sparse Superposition Hypothesis You can encode n-dimensional k-sparse vectors in a much lower than n-dimensional space. If you think neural networks...

Comment from Yasaman Bahri (@yasamanb): maybe the reason we see poly-semantic neurons is that the task isn't hard enough to get neurons in later layers to learn the "right" abstractions....

I did a hacky version of this for html (https://github.com/tensorflow/lucid/blob/master/lucid/misc/io/saving.py#L184). Unfortunately, my version slows down save a bunch because it separately invokes gsutil. I think @ludwigschubert has an idea for...