© 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. 15459
    Prev
    Next
    INDEX
    Explanations

    text layout, columns, table headers

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

    This neuron fires strongly on unusually informative or low-frequency content tokens (i.e. rare “content” words and numerals).

    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
    a
    0.86
     serez
    0.84
     Include
    0.76
    
    0.76
    س
    0.74
     NEX
    0.73
    s
    0.73
    芙
    0.73
     voulez
    0.72
     Miley
    0.72
    POSITIVE LOGITS
     ligados
    0.95
     паралле
    0.84
     первые
    0.82
     emphasizing
    0.81
     полно
    0.79
     odors
    0.79
     нужны
    0.77
     друзья
    0.77
     capazes
    0.77
    DropdownBox
    0.76
    Activations Density 0.000%

    No Known Activations