© 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. 102249
    Prev
    Next
    INDEX
    Explanations

    ask me for something

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

    The neuron detects imperative commands or requests that use “me” (e.g. “call me,” “help me,” “show me”).

    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
    つも
    -1.72
     ceket
    -1.69
     entstand
    -1.66
     conosco
    -1.63
    ſelves
    -1.62
     också
    -1.61
     PUEDE
    -1.57
     åren
    -1.57
     violento
    -1.56
     milj
    -1.54
    POSITIVE LOGITS
     These
    1.98
     It
    1.85
     Each
    1.68
     This
    1.65
     in
    1.63
     Whether
    1.56
     by
    1.55
     if
    1.50
     Since
    1.48
     If
    1.46
    Activations Density 0.029%

    No Known Activations