© 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. 15114
    Prev
    Next
    INDEX
    Explanations

    The neuron fires most strongly on tokens containing digits or special symbols (e.g. numbers, punctuation marks, error‐codes, and the question mark).

    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
    ان
    0.67
     scriptures
    0.66
    an
    0.61
    iston
    0.60
    न
    0.58
    quele
    0.57
    が必要です
    0.56
     желательно
    0.56
    を行
    0.55
     здраво
    0.55
    POSITIVE LOGITS
     keď
    0.66
     ঘটেছে
    0.61
     když
    0.59
     Kategorie
    0.58
    ກັບ
    0.56
    (
    0.55
    𝐬
    0.55
     misalnya
    0.55
     mecz
    0.55
     vš
    0.54
    Activations Density 16.896%

    No Known Activations