© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 38541
    Prev
    Next
    INDEX
    Explanations

    is really

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

    The neuron selectively activates on numeric tokens (including integers and decimals), effectively spotting any numbers or numerical expressions in the text.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
    すっ
    -1.23
    ほら
    -1.18
     berpeng
    -1.13
     mengajak
    -1.07
    ようです
    -1.07
    volve
    -1.06
    коро
    -1.05
    凄く
    -1.05
    明日は
    -1.04
    為に
    -1.04
    POSITIVE LOGITS
     is
    2.48
     was
    2.19
     are
    1.79
     seem
    1.66
     seemed
    1.50
     really
    1.48
     especially
    1.43
     seems
    1.36
     could
    1.33
     I
    1.30
    Activations Density 0.043%

    No Known Activations