© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 127857
    Prev
    Next
    INDEX
    Explanations

    The neuron is picking out very high-frequency tokens—especially standalone numbers and common function words.

    oai_token-act-pair · o4-mini
    Triggered by @jyhe0408

    referred to

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Negative Logits
     Außerdem
    -0.89
    Although
    -0.86
     Although
    -0.84
     hilsen
    -0.84
     bereits
    -0.83
    ᎒
    -0.82
    PhoneNumber
    -0.81
     Since
    -0.80
    StringBuffer
    -0.78
     snart
    -0.78
    POSITIVE LOGITS
     then
    0.94
    ucing
    0.89
    sendLogPayload
    0.89
     lã
    0.89
    ாட
    0.88
    liances
    0.87
    jonal
    0.86
     ricerche
    0.85
     filós
    0.85
     вами
    0.84
    Activations Density 0.010%

    No Known Activations