© 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. 84531
    Prev
    Next
    INDEX
    Explanations

    pard, amnesty, Q, Millions

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

    This neuron flags capitalized tokens that are part of proper names or organizations.

    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
    咘
    -2.45
    不
    -2.39
    甃
    -2.28
     We
    -2.23
     bytu
    -2.23
    神
    -2.22
    箧
    -2.22
    炱
    -2.19
     tisíc
    -2.19
    </h3>
    -2.19
    POSITIVE LOGITS
    ↵↵
    3.80
    ↵
    2.91
    5
    2.86
    7
    2.69
    We
    2.61
    <i>
    2.50
    4
    2.44
    ’
    2.34
    ;
    2.33
    </b>
    2.31
    Activations Density 0.002%

    No Known Activations