© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 129218
    Prev
    Next
    INDEX
    Explanations

    The main thing this neuron does is pick out numerical and numeric‐related tokens (like literal digits, numeric constants, and comparison or assignment operators) across code and technical descriptions.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    decided, change, reaction

    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
    Not in Any Lists

    No Comments

    Negative Logits
     any
    -1.16
    分ほど
    -0.93
    GUILayout
    -0.89
     Unwin
    -0.89
    UEZ
    -0.87
    ">\
    -0.85
    featureID
    -0.84
    inline
    -0.84
    ￣
    -0.82
    рения
    -0.81
    POSITIVE LOGITS
     Kemudian
    0.89
    曲は
    0.83
    exels
    0.81
     イエロー
    0.81
     Aviso
    0.81
     otros
    0.81
     japones
    0.79
    komendasi
    0.79
     nace
    0.77
    रूप
    0.77
    Activations Density 0.092%

    No Known Activations