© 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. 3827
    Prev
    Next
    INDEX
    Explanations

    The neuron strongly fires on the standalone numerals that mark the start of each numbered reference (i.e. the leading citation numbers in a bibliography).

    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
    él
    0.54
    朝着
    0.52
    be
    0.52
    选择了
    0.52
    ────────────────
    0.52
    one
    0.51
    ↵
    0.50
    ého
    0.50
    :
    0.50
    anger
    0.50
    POSITIVE LOGITS
     Maintains
    0.63
    VhZ
    0.63
     Turquoise
    0.62
     IComparable
    0.61
     Daarna
    0.61
     Graduates
    0.61
     Payroll
    0.59
    Istituto
    0.59
     MDLVertex
    0.59
     VEGFR
    0.59
    Activations Density 0.000%

    No Known Activations