© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 86624
    Prev
    Next
    INDEX
    Explanations

    Damien Brunner, Stamford Bridge, Fallon Fox

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

    The neuron chiefly activates on proper-noun tokens—names of people, places, organizations, and other capitalized entities.

    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
    \
    -3.48
    桕
    -2.58
    9
    -2.34
     animosity
    -2.28
    7
    -2.27
    <i>
    -2.25
    ﺅ
    -2.25
    *
    -2.22
     Those
    -2.19
    4
    -2.19
    POSITIVE LOGITS
    裢
    3.34
    )
    3.11
    蛲
    2.86
    鸸
    2.78
     afirmó
    2.67
    actéristi
    2.58
     expresó
    2.53
     usu
    2.53
    蹰
    2.53
    ᯾
    2.50
    Activations Density 0.011%

    No Known Activations