© 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. 22-GEMMASCOPE-TRANSCODER-16K
    7. 13230
    Prev
    Next
    INDEX
    Explanations

    words starting with "cul" or "mul," with a slightly stronger preference for "cul."

    oai_token-act-pair · gemini-2.0-flash

    Prefix-like word parts

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2b-pt-transcoders/layer_22/width_16k/average_l0_15
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.22.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
     bedeckt
    -0.56
    XmlAccessType
    -0.53
    ilers
    -0.52
    CodedInputStream
    -0.49
     snippetHide
    -0.49
    unks
    -0.49
    nthesis
    -0.48
    atabase
    -0.48
    orks
    -0.47
    uarts
    -0.47
    POSITIVE LOGITS
    ::_('
    0.60
    let
    0.55
    "});
    0.55
     EFI
    0.54
     prossima
    0.53
    لكتر
    0.53
    principalTable
    0.52
     externi
    0.52
    __);
    0.51
     gostar
    0.51
    Activations Density 0.609%

    No Known Activations