© 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. 34-GEMMASCOPE-RES-131K
    7. 76414
    Prev
    Next
    INDEX
    Explanations

    places starting with kah or koh

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

    The neuron consistently fires on short tokens beginning with the letter “K” immediately followed by a single vowel (e.g. “Ka,” “Ke,” “Ki,” “Ko,” “Ku”).

    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
    ViewComponent
    -0.84
     навіть
    -0.75
     even
    -0.75
    Цитата
    -0.74
    こんばんは
    -0.73
     substantially
    -0.72
    ാല
    -0.72
     beyaz
    -0.72
     Additional
    -0.72
     飲
    -0.71
    POSITIVE LOGITS
     mods
    0.80
    frac
    0.79
    indo
    0.77
    mediated
    0.73
     crashes
    0.73
    招
    0.73
    keres
    0.71
     XVII
    0.70
     Beratung
    0.69
    fran
    0.69
    Activations Density 0.017%

    No Known Activations