© 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. 755
    Prev
    Next
    INDEX
    Explanations

    The neuron selectively activates on floating‐point numerical tokens (decimal numbers, often year‐like values).

    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
     new
    0.63
     mio
    0.62
    Name
    0.61
    '])[
    0.58
    0.57
     vostro
    0.53
    <0xE3>
    0.52
     more
    0.52
     a
    0.51
     X
    0.51
    POSITIVE LOGITS
     upbringing
    0.83
    经历了
    0.80
     filmmaking
    0.77
    experiences
    0.71
    经历
    0.69
     прошла
    0.68
     στη
    0.67
     बतौर
    0.66
    身为
    0.66
     stints
    0.65
    Activations Density 0.017%

    No Known Activations