© 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. 31-GEMMASCOPE-2-RES-16K
    7. 1692
    Prev
    Next
    INDEX
    Explanations

    in followed by punctuation

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

    The neuron strongly activates on marketing-style or promotional words (“free,” “choose,” “more,” “every,” “kind,” etc.), i.e. terms commonly used in calls-to-action or advertising copy.

    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_31_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
    ClassLoader
    0.56
     анд
    0.51
     Irishman
    0.51
    спеди
    0.49
    ảy
    0.48
    ℍ
    0.47
    לו
    0.47
    לב
    0.47
     lượt
    0.46
     términos
    0.46
    POSITIVE LOGITS
     another
    0.58
    zoa
    0.55
    another
    0.54
    entities
    0.54
     prettier
    0.52
    grand
    0.52
    licensed
    0.52
     مذہبی
    0.52
    饬
    0.52
    ients
    0.51
    Activations Density 0.003%

    No Known Activations