© 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. 43669
    Prev
    Next
    INDEX
    Explanations

    best friend

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

    The neuron activates strongly on occurrences of the word “friend,” especially in the phrase “best friend.”

    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
    柽
    -1.91
    旆
    -1.70
    菝
    -1.66
    ࿊
    -1.60
     maxim
    -1.59
    pirate
    -1.56
    maquilla
    -1.54
    倣
    -1.51
    Infant
    -1.50
    塬
    -1.50
    POSITIVE LOGITS
    ORROW
    1.67
    俢
    1.62
    次に
    1.61
    それを
    1.60
    //
    1.55
     militer
    1.53
     Очень
    1.52
     backen
    1.52
     малень
    1.52
    ной
    1.52
    Activations Density 0.007%

    No Known Activations