© 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. 2340
    Prev
    Next
    INDEX
    Explanations

    sentence end punctuation

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

    The neuron fires on verbs and punctuation that introduce or tag quoted speech (e.g. “said,” “says,” “added,” “recalls,” etc.), i.e. speech‐attribution contexts.

    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
     मेरे
    0.70
    こういう
    0.62
     М
    0.60
     stupid
    0.60
     इस
    0.59
     grievous
    0.59
     এই
    0.58
     flimsy
    0.57
    А
    0.57
     А
    0.56
    POSITIVE LOGITS
    olo
    0.61
    ície
    0.59
    τής
    0.58
    ικη
    0.58
    acido
    0.58
    miştir
    0.57
    ிறார்
    0.57
    lán
    0.57
    づくり
    0.57
     zunehm
    0.57
    Activations Density 0.080%

    No Known Activations