© 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. 20-GEMMASCOPE-TRANSCODER-16K
    7. 15740
    Prev
    Next
    INDEX
    Explanations

    seemingly unrelated, the neuron is most likely looking for names, specifically Arabic names

    oai_token-act-pair · gemini-2.0-flash

    ī

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2b-pt-transcoders/layer_20/width_16k/average_l0_11
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.20.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
    httphttps
    -0.81
    tvguidetime
    -0.79
     متعلقه
    -0.78
    FunctionFlags
    -0.74
    parsedMessage
    -0.69
    ⓧ
    -0.67
     StatefulWidget
    -0.66
    CWE
    -0.65
     ostavi
    -0.63
     للمعارف
    -0.61
    POSITIVE LOGITS
     Paglinawan
    0.44
     perpetuity
    0.44
     natin
    0.42
    expires
    0.39
     negó
    0.39
     méta
    0.39
    といけない
    0.39
     싶
    0.38
     yüzden
    0.38
    systemctl
    0.37
    Activations Density 0.244%

    No Known Activations