© 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-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 102649
    Prev
    Next
    INDEX
    Explanations

    measurements or quantities

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

    The neuron responds to the digit “5” when it appears as part of numeric measurements—especially in decimal fractions ending in “.5.”

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    {
    -2.19
     sū
    -1.98
     iſt
    -1.92
     zimowe
    -1.90
    ⢌
    -1.81
     tokyo
    -1.80
    ㌘
    -1.66
     kimdir
    -1.65
     dezelve
    -1.63
    '^
    -1.62
    POSITIVE LOGITS
    {~
    2.19
     rheinland
    2.06
    when
    2.06
    this
    2.00
    }{
    1.80
    </sup>
    1.79
     braucht
    1.73
    chtenstein
    1.70
     brauche
    1.69
     ewigen
    1.69
    Activations Density 0.013%

    No Known Activations