© 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. 22-GEMMASCOPE-RES-131K
    7. 92879
    Prev
    Next
    INDEX
    Explanations

    names and titles

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

    This neuron fires on document‐header metadata tokens—e.g. the all‐caps place names, numbers, codes, dates, and other uppercase or numeric strings that appear in datelines, bylines, headings, and URLs.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
    <eos>
    -1.52
    ά
    -0.94
     decimos
    -0.92
    versary
    -0.87
    QUIT
    -0.85
    Quellen
    -0.85
    cales
    -0.84
    sa
    -0.83
     bekam
    -0.83
    فعل
    -0.83
    POSITIVE LOGITS
     wierd
    0.88
    یح
    0.85
    pondi
    0.84
    dchen
    0.83
     sapin
    0.81
     coincided
    0.81
    *$
    0.79
     soloist
    0.78
     prenn
    0.78
     tahan
    0.77
    Activations Density 0.002%

    No Known Activations