© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 8-RES-JB
    11. 2215
    Prev
    Next
    INDEX
    Explanations

    ingredients and recipe instructions for cooking

    oai_token-act-pair · gpt-3.5-turbo

    culinary instructions and ingredients used in recipes

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.hook_resid_pre
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    •single token firings
    Fires individually on " tablespoons", " skillet", " cloves", " oven" and lots of ingredients.

    No Comments

    Negative Logits
    href
    -0.61
    pheus
    -0.58
     tacit
    -0.57
     headlined
    -0.57
     Fundamental
    -0.56
     patrols
    -0.56
     insider
    -0.56
     exploits
    -0.55
    ¥µ
    -0.55
    Corp
    -0.54
    POSITIVE LOGITS
    Ingredients
    1.05
     yeast
    1.00
     Ingredients
    1.00
     roasted
    0.95
     fermentation
    0.92
     baking
    0.89
     recipe
    0.87
     caul
    0.87
     ingredients
    0.87
     broth
    0.87
    Activations Density 0.522%

    No Known Activations