© 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. 34-GEMMASCOPE-RES-131K
    7. 84071
    Prev
    Next
    INDEX
    Explanations

    names ending in 'oe' or 'e'

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

    This neuron strongly activates on capitalized proper names or technical terms—i.e. tokens representing named 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_34/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
     contado
    -0.89
    縋
    -0.82
     resultant
    -0.82
    bited
    -0.82
    の場合は
    -0.81
     légère
    -0.80
     подвод
    -0.79
     immédi
    -0.79
    stt
    -0.79
    kalender
    -0.77
    POSITIVE LOGITS
     اصل
    0.76
     zeit
    0.72
    キーワード
    0.71
    究極
    0.71
    ètement
    0.71
     dera
    0.70
    ovali
    0.69
    tagName
    0.68
     tuba
    0.68
    "}>
    0.68
    Activations Density 0.019%

    No Known Activations