© 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. 1052
    Prev
    Next
    INDEX
    Explanations

    The neuron strongly activates on the initial sub‐tokens of proper names and technical terms (i.e. the leading capitalized pieces of named entities).

    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
    ่
    0.57
     einzelnen
    0.55
    RE
    0.55
     ದೇಶ
    0.55
    jej
    0.54
    aaa
    0.52
    viso
    0.52
    ično
    0.52
     اما
    0.51
    눙
    0.51
    POSITIVE LOGITS
    ists
    0.62
    etary
    0.61
    ization
    0.61
    ogical
    0.60
    ological
    0.59
    ically
    0.58
    ları
    0.57
    ifying
    0.56
    ily
    0.55
    osphere
    0.55
    Activations Density 0.000%

    No Known Activations