© 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. 6369
    Prev
    Next
    INDEX
    Explanations

    This neuron strongly activates on numeric tokens and symbols (e.g. digits, parentheses around numbers, version numbers), i.e. it detects numerical information.

    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
    relative
    0.80
     በአ
    0.75
     pourcentage
    0.73
    shooting
    0.72
     qualche
    0.71
     протягом
    0.71
     በ
    0.71
    discover
    0.69
     Secondly
    0.68
     размера
    0.68
    POSITIVE LOGITS
     humanos
    0.82
    дая
    0.79
     flagship
    0.76
     judíos
    0.75
    াণ
    0.75
     bú
    0.74
     нужны
    0.73
     diploid
    0.71
    रियो
    0.71
    轶
    0.71
    Activations Density 0.000%

    No Known Activations