© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 104749
    Prev
    Next
    INDEX
    Explanations

    how can i/you/we

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

    This neuron detects questions phrased with modal auxiliaries (especially “can I/you,” “how can,” or “could I”) at the start of a sentence.

    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
    Podob
    -2.02
    Makanan
    -1.74
    Teraz
    -1.66
    Če
    -1.65
     手绘
    -1.63
    Якщо
    -1.63
    Selamat
    -1.61
     tradicionais
    -1.60
     土産
    -1.60
    因為
    -1.58
    POSITIVE LOGITS
     which
    1.93
    ↵↵
    1.91
     so
    1.77
     не
    1.57
     com
    1.50
     information
    1.48
     that
    1.42
    com
    1.41
     I
    1.40
    не
    1.40
    Activations Density 0.137%

    No Known Activations