© 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-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 6978
    Prev
    Next
    INDEX
    Explanations

    calculates

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

    The neuron fires on action verbs in the simple present tense (including imperatives and third-person singular forms) that drive calls to action or describe product features.

    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
     leveraged
    0.82
    чны
    0.74
     INTERACTIONS
    0.72
     stakeholders
    0.71
    ร่วม
    0.67
    นิยม
    0.67
    rgb
    0.66
    認
    0.65
    個別
    0.65
    новить
    0.64
    POSITIVE LOGITS
    ಾಗಲೇ
    0.77
    اً
    0.74
    اس
    0.72
    ส์
    0.72
     、,
    0.71
     excitedly
    0.68
     Untuk
    0.67
     Postage
    0.67
    verture
    0.67
    stitution
    0.67
    Activations Density 0.002%

    No Known Activations