Neuronpedia API Reference
IMPORTANT: Check the available resources page to see which models and sources are available in the API. API KEY: Get your API key by logging in at neuronpedia.org then going to /account. Requests for new APIs, questions, or feedback? Contact us. Gets activation values for all features in a source/SAE when processing custom input text(s). It returns the activations and tokenized prompts only, without any of the dashboards or explanations. The custom text to process. Either a single string or an array of strings. If it's a single string, the max is a 1024 token text. If it's an array, the max is 4 strings of 256 tokens max each. The custom text to process. Either a single string or an array of strings. If it's a single string, the max is a 1024 token text. If it's an array, the max is 4 strings of 256 tokens max each. The model the source is in. The source/SAE ID. Successful response, with activation data Bad request, missing or invalid input Unauthorized, user doesn't have access to the