© 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. 10577
    Prev
    Next
    INDEX
    Explanations

    text content

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

    The neuron selectively activates on document section headings or labels (e.g. “Personal Narrative,” “Description,” “Content,” “Format”) that mark structural or topical breaks.

    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
    contador
    0.69
    IED
    0.63
     لاعب
    0.61
    ZC
    0.61
    eig
    0.59
    AE
    0.57
    伫
    0.57
    玩家
    0.57
    गराज
    0.57
    ounter
    0.56
    POSITIVE LOGITS
     QText
    0.88
     setText
    0.86
     text
    0.83
     texte
    0.83
     textos
    0.80
     текст
    0.79
     내용
    0.77
     texto
    0.76
     texts
    0.75
     textarea
    0.73
    Activations Density 3.153%

    No Known Activations