© 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. 47966
    Prev
    Next
    INDEX
    Explanations

    king/queen of titles

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

    The neuron strongly responds to the preposition “of” in two-part title or rank constructions (e.g. “King of X,” “Queen of Y,” “Prince of Z”).

    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
     a
    -2.66
    ?
    -2.03
    ↵↵
    -1.96
     $
    -1.88
    2
    -1.80
    .”
    -1.78
    .
    -1.74
    <bos>
    -1.65
     '
    -1.59
    \
    -1.59
    POSITIVE LOGITS
    cesor
    1.84
    媜
    1.83
    ordinator
    1.83
     immagini
    1.82
    henswürdigkeiten
    1.82
     FÜR
    1.78
    arschu
    1.73
    Това
    1.72
    硨
    1.71
     However
    1.69
    Activations Density 0.017%

    No Known Activations