© 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-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 9291
    Prev
    Next
    INDEX
    Explanations

    action words

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

    The neuron strongly activates on lexical “content” verbs that describe main actions (e.g. watched, listened, wrote, read).

    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
    Се
    0.62
    {
    0.61
    несла
    0.60
    ת
    0.60
     Međutim
    0.59
    公開
    0.58
    Бу
    0.58
    Ку
    0.57
     ganhou
    0.57
    꿩
    0.57
    POSITIVE LOGITS
     וא
    0.60
     gill
    0.59
     bepaalde
    0.59
     perils
    0.59
     cartas
    0.58
     comandos
    0.57
     préférences
    0.57
     SDGs
    0.57
    眶
    0.57
     seçim
    0.56
    Activations Density 0.001%

    No Known Activations