© 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. 10-GEMMASCOPE-RES-131K
    7. 84619
    Prev
    Next
    INDEX
    Explanations

    titles and noble names

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

    The neuron strongly activates on tokens that are parts of proper names—especially aristocratic family names and noble titles.

    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
     古い
    -1.38
     全新
    -1.38
    Цита
    -1.36
     nyní
    -1.34
    продам
    -1.34
    petito
    -1.30
     나무
    -1.30
    ;
    -1.27
    }
    -1.27
    '
    -1.24
    POSITIVE LOGITS
    ogues
    1.82
    datable
    1.81
    ???:
    1.66
    strator
    1.66
    naires
    1.65
    omato
    1.64
    gered
    1.60
    zies
    1.60
    yled
    1.60
    corer
    1.59
    Activations Density 0.056%

    No Known Activations