Files with activations to run viewer locally

[WuTheFWasThat]

[this issue is migrated from a previous version of the repo]

@elephantmipt and @zdaiot ask:

Hi, thank you for releasing great tool! It would be great if you provide more details on which files SAE viewer uses for rendering particular feature. When I run published code locally, I get an error Failed to load resource: net::ERR_CONNECTION_REFUSED because here

sparse_autoencoder/sae-viewer/src/interpAPI.ts

Lines 4 to 20 in 0296e02

	export const load_file_no_cache = async(path: string) => {
	const data = {
	path: path
	}
	const url = new URL("/load_az", window.location.href)
	url.port = '8000';
	return await (
	await fetch(url, {
	method: "POST", // or 'PUT'
	headers: {
	"Content-Type": "application/json",
	},
	body: JSON.stringify(data),
	})
	).json()
	}

we try to connect to localhost:8000 load_az method.

I need an example of a file with activations that I can pass as JSON here

sparse_autoencoder/sae-viewer/src/interpAPI.ts

Line 55 in 0296e02

const result = await load_fn(`${prefix}/${ae.subject}/${ae.path}/atoms/${feature.atom}${ablated ? '-ablated': ''}.json`)

, in case I want to view features from my custom SAE. However, I am unsure of the format to use. Could you please provide example files with activations so that anyone can run the demo locally and/or modify the activations and models?

[WuTheFWasThat]

example file: https://openaipublic.blob.core.windows.net/sparse-autoencoder/viewer/gpt2-small/v5_32k/layer_8/resid_post_mlp/atoms/0-ablated.json

the ablation data is optional, simpler format is:

https://openaipublic.blob.core.windows.net/sparse-autoencoder/viewer/gpt2-small/v5_32k/layer_8/resid_post_mlp/atoms/0.json

[zdaiot]

@WuTheFWasThat Can you give me the process of how to generate these json files?

[WuTheFWasThat]

we just ran the model+autoencoder over a pretraining dataset and tracked highest activations per latent and random positive activations per latent. we don't have a plan to release the code atm, but if that would be helpful we can consider it

[zdaiot]

we just ran the model+autoencoder over a pretraining dataset and tracked highest activations per latent and random positive activations per latent. we don't have a plan to release the code atm, but if that would be helpful we can consider it

Thanks a lot. I think it will be very helpful. Please consider it.

Thank you for such an excellent job. Can I ask, when release training code?

[taha-yassine]

Hi,
Thanks for the excellent work!
l'm also interested in the code to generate the JSONs if you ever decide to release. I think it would be helpful.

[w1051868626]

Me too。I'm also interested in generating JSON code if you decide to publish it. I think it'll help.