© 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. 13626
    Prev
    Next
    INDEX
    Explanations

    train, belonging, playing

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

    This neuron strongly activates on tokens that occur at the ends of sentences (i.e. sentence-final words).

    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
    on
    1.01
    a
    0.99
    ai
    0.90
     Lur
    0.88
    ^{-
    0.79
     Cutting
    0.78
     Carrying
    0.78
     Checking
    0.76
     Duties
    0.75
     Disability
    0.75
    POSITIVE LOGITS
     rougeâtres
    0.84
    結束
    0.77
    Με
    0.77
    ം
    0.75
    最終
    0.75
    પ
    0.74
     élargies
    0.73
     юго
    0.73
    Estamos
    0.73
    最终
    0.71
    Activations Density 0.000%

    No Known Activations