© 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. 0-RES-JB
    11. 1817
    Prev
    Next
    INDEX
    Explanations

    the name "Gan" or mentions of it

    oai_token-act-pair · gpt-3.5-turbo

    variations of the word "gan" (short for "vegetarian" or "vegan")

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     TTL
    -0.79
     ante
    -0.69
     tables
    -0.64
     Seat
    -0.63
     exhibits
    -0.63
     TR
    -0.62
     TL
    -0.61
     losers
    -0.61
     packages
    -0.60
     Hyde
    -0.60
    POSITIVE LOGITS
    gan
    4.66
    gans
    2.20
    GAN
    1.71
    ghan
    1.69
    gins
    1.62
    gin
    1.59
    agan
    1.31
    gyn
    1.26
    gran
    1.25
    aghan
    1.25
    Activations Density 0.012%

    No Known Activations