© 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
    2. Gemma-3-12B
    3. 24-GEMMASCOPE-2-RES-16K
    4. 1458
    Prev
    Next
    INDEX
    Explanations

    The neuron responds to low‐frequency or unusual tokens (e.g. rare proper names, the “@” symbol in email addresses, unique capitalized words like “Grace,” etc.).

    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
     sighted
    0.71
    B
    0.70
     vi
    0.68
     GE
    0.65
    :
    0.64
    GE
    0.63
     ش
    0.63
     CE
    0.63
     dot
    0.62
    tower
    0.61
    POSITIVE LOGITS
    रोग
    1.01
     Filt
    0.95
     Produk
    0.94
    bahan
    0.89
     psik
    0.89
     kebutuhan
    0.88
     необходимости
    0.87
     necesitamos
    0.86
     usia
    0.86
     kontrol
    0.86
    Activations Density 0.000%

    No Known Activations