© 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. 31-GEMMASCOPE-2-RES-16K
    7. 1844
    Prev
    Next
    INDEX
    Explanations

    or

    np_acts-logits-general · gemini-2.5-flash-lite

    The neuron is triggered by standalone numerical tokens, especially floating-point numbers (i.e. decimal measurements).

    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_31_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
    <blockquote>
    1.61
    According
    1.61
    While
    1.59
    Despite
    1.57
    <h3>
    1.52
    The
    1.51
    Although
    1.50
    In
    1.47
    Since
    1.47
    <h4>
    1.46
    POSITIVE LOGITS
     analyte
    0.95
     nucleons
    0.88
     alkyl
    0.88
     covariance
    0.88
     homogene
    0.83
     glucos
    0.81
     essere
    0.80
     monomial
    0.80
     subarray
    0.80
     tipo
    0.79
    Activations Density 0.455%

    No Known Activations