© 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. 10-GEMMASCOPE-RES-131K
    7. 100831
    Prev
    Next
    INDEX
    Explanations

    division operations

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

    the word "divided" followed by "by" in mathematical division problems.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @bfgcn7frf5

    The neuron fires on tokens marking a division operation (e.g. “div,” “divided,” “divide by”).

    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
    P
    -1.88
    S
    -1.87
    J
    -1.66
    なども
    -1.62
    i
    -1.62
    V
    -1.58
    N
    -1.57
    K
    -1.55
     таке
    -1.55
    琑
    -1.55
    POSITIVE LOGITS
     "
    1.81
     modernas
    1.75
    個人的
    1.74
    ほら
    1.70
     Çok
    1.66
    Karakteristik
    1.66
    Sommaire
    1.57
    Responsable
    1.56
    danke
    1.55
     incurring
    1.54
    Activations Density 0.005%

    No Known Activations