© 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. 22-GEMMASCOPE-RES-131K
    7. 93024
    Prev
    Next
    INDEX
    Explanations

    questions starting with "what" or "how"

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

    The neuron activates strongly on interrogative question words (e.g. What, Who, How, Why) at the start of questions.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
     acheteurs
    -1.44
     consommateurs
    -1.27
     trajets
    -1.27
    ølge
    -1.22
     fournisseurs
    -1.21
     honden
    -1.20
    ğan
    -1.20
     navires
    -1.19
    rée
    -1.18
     Tatsache
    -1.16
    POSITIVE LOGITS
     then
    2.23
     other
    1.83
     to
    1.80
     different
    1.47
     so
    1.45
     you
    1.37
     now
    1.32
     about
    1.31
     does
    1.31
     więc
    1.28
    Activations Density 0.014%

    No Known Activations