I couldn’t find a link to the code in the article so in case anyone else wants to try to replicate I think this is it: https://github.com/HugoFry/mats_sae_training_for_ViTs
Louka Ewington-Pitsos
Food, Prison & Exotic Animals: Sparse Autoencoders Detect 6.5x Performing Youtube Thumbnails
Massive Activations and why <bos> is important in Tokenized SAE Unigrams
Training a Sparse Autoencoder in < 30 minutes on 16GB of VRAM using an S3 cache
Just to close the loop on this one, the official huggingface transformers library just uses a for-loop to achieve MoE. I also implemented a version myself using a for loop and it’s much more efficient than either vanilla matrix multiplication or that weird batch matmul I write up there for large latent and batch sizes.
wait a minute… could you just...
you don’t just literally do this do you?
input = torch.tensor([ [1, 2], [1, 2], [1, 2], ]) # (bs, input_dim) enc_expert_1 = torch.tensor([ [1, 1, 1, 1], [1, 1, 1, 1], ]) enc_expert_2 = torch.tensor([ [3, 3, 0, 0], [0, 0, 2, 0], ]) dec_expert_1 = torch.tensor([ [ -1, -1], [ -1, -1], [ -1, -1], [ -1, -1], ]) dec_expert_2 = torch.tensor([ [-10, -10,], [-10, -10,], [-10, -10,], [-10, -10,], ]) def moe(input, enc, dec, nonlinearity): input = input.unsqueeze(1) latent = torch.bmm(input, enc) recon = torch.bmm(nonlinearity(latent, dec)) return recon.squeeze(1), latent.squeeze(1) # not this! some kind of actual routing algorithm, but you end up with something similar enc = torch.stack([enc_expert_1, enc_expert_2, enc_expert_1]) dec = torch.stack([dec_expert_1, dec_expert_2, dec_expert_1]) nonlinearity = torch.nn.ReLU() recons, latent = moe(input, enc, dec, nonlinearity)
This must in some way be horrifically inefficient, right?
Can I ask what you used to implement the MOE routing? Did you use megablocks? I would love to expand on this research but I can’t find any straightforward implementation of efficient pytorch MOE routing online.
Do you simply iterate over each max probability expert every time you feed in a batch?
Faithful vs Interpretable Sparse Autoencoder Evals
This is dope, thank you for your service. Also, can you hit us with your code on this one? Would love to reproduce.
It seems like it’s possible right?
Interestingly a similar tool run by a very smart guy (https://www.creatorml.com/) is apparently about to shut down, so it might not be a financially sustainable thing to try.