© 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. 102948
    Prev
    Next
    INDEX
    Explanations

    Grady, Longo, ADO, AIAA, LBL, LABEL

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

    The neuron selectively activates on the proper name “Grady” (e.g. in “Grady W. Brown,” “Camp Grady,” “Minister Grady”).

    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
    饹
    -2.91
    
    -2.78
    
    -2.72
    乆
    -2.53
    枼
    -2.41
    沨
    -2.41
    烀
    -2.30
    訫
    -2.30
    ⟞
    -2.27
    ve
    -2.25
    POSITIVE LOGITS
    占い
    2.08
     ");
    1.90
    有着
    1.87
     demonstrates
    1.86
     они
    1.86
     fortsetter
    1.81
     cbd
    1.80
     готовы
    1.80
     ikkje
    1.79
    鼉
    1.79
    Activations Density 0.001%

    No Known Activations