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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 70902
    Prev
    Next
    INDEX
    Explanations

    act of writing

    np_acts-logits-general · gemini-2.5-flash-lite

    This neuron fires on mentions of writing or implementing code (e.g. “write,” “wrote,” “writing”) usually in the context of functions or programs.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
    ’
    -1.84
    ignal
    -1.49
     ustal
    -1.45
     péché
    -1.33
     dì
    -1.33
     Kopi
    -1.33
    {
    -1.32
    ndo
    -1.31
     Couleur
    -1.30
     rythme
    -1.28
    POSITIVE LOGITS
     This
    1.49
    Biographies
    1.45
    osene
    1.45
    はその
    1.38
    缛
    1.38
     wrote
    1.35
     написать
    1.32
    derdag
    1.30
    LEGGI
    1.29
    偸
    1.29
    Activations Density 0.098%

    No Known Activations