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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-2B
    5. 21-GEMMASCOPE-TRANSCODER-16K
    7. 12955
    Prev
    Next
    INDEX
    Explanations

    words that are often used by professionals and academics

    oai_token-act-pair · gemini-2.0-flash

    generic article words

    np_token-act-pair-logits · gpt-4o-mini

    Start of phrases

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2b-pt-transcoders/layer_21/width_16k/average_l0_13
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.21.ln2.hook_normalized
    Architecture
    jumprelu_transcoder
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Majefty
    -0.76
     Monfieur
    -0.72
     Efq
    -0.72
     Jefus
    -0.71
     myſelf
    -0.69
     itſelf
    -0.69
     Chriftian
    -0.67
     Shakspeare
    -0.65
     poffible
    -0.65
     themſelves
    -0.65
    POSITIVE LOGITS
     the
    0.78
     this
    0.74
     her
    0.71
     these
    0.71
     one
    0.70
     pres
    0.70
     for
    0.69
     is
    0.69
    __:
    0.68
     his
    0.68
    Activations Density 7.986%

    No Known Activations