© 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. 100480
    Prev
    Next
    INDEX
    Explanations

    at risk of

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

    The neuron selectively activates on numeric tokens—integers or decimals—wherever quantities, percentages, or other numbers appear.

    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
    '
    -1.83
    -
    -1.74
    ,'
    -1.66
    ',
    -1.63
     ('
    -1.62
    ↵
    -1.54
    .'
    -1.52
     gier
    -1.46
    ",
    -1.45
     cepillo
    -1.44
    POSITIVE LOGITS
     medes
    1.69
     competen
    1.60
     as
    1.57
     simpat
    1.52
     by
    1.51
     same
    1.50
    EDEFAULT
    1.47
     luxo
    1.45
     decora
    1.44
     destra
    1.40
    Activations Density 0.181%

    No Known Activations