© 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. 34-GEMMASCOPE-RES-131K
    7. 93402
    Prev
    Next
    INDEX
    Explanations

    Can I questions

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

    The neuron fires on polite first‐person requests or questions—spotting words like “I,” “please,” and modal verbs (“Can,” “May,” etc.) that mark someone asking for help or permission.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
    ្រ
    -1.07
     allowing
    -1.02
    Need
    -1.01
     probabilmente
    -0.95
     rekening
    -0.94
     moeten
    -0.94
     needed
    -0.93
     probably
    -0.92
     necessidade
    -0.92
     want
    -0.92
    POSITIVE LOGITS
     please
    2.11
     PLEASE
    1.59
    please
    1.47
     Please
    1.28
     пожалуйста
    1.26
    Please
    1.23
     get
    1.23
     interest
    1.21
    PLEASE
    1.09
     use
    1.08
    Activations Density 0.064%

    No Known Activations