© 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
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 12928
    Prev
    Next
    INDEX
    Explanations

    code setup or definition

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

    The neuron strongly activates on code comments phrased as imperative instructions (e.g. “Keep track…,” “Select this instruction,” “Dump the range…,” “Minimize Altia Runtime…”), i.e. directive or TODO‐style comments.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_24_width_16k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    <strong>
    0.94
    <b>
    0.83
    L
    0.83
    :
    0.81
    Type
    0.78
    You
    0.77
    《
    0.70
    Fig
    0.69
    Г
    0.69
    V
    0.68
    POSITIVE LOGITS
     atuais
    1.02
     asesinato
    1.02
     aktuellen
    0.95
    зываем
    0.93
     keinginan
    0.93
     Estamos
    0.92
     überprüfen
    0.91
     стрельца
    0.91
    roffenen
    0.90
     เด้อ
    0.89
    Activations Density 0.004%

    No Known Activations