© 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. 22-GEMMASCOPE-RES-131K
    7. 34416
    Prev
    Next
    INDEX
    Explanations

    noteworthy aspects

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

    This neuron activates on words and phrases that convey emphasis or high importance—such as superlative or intensifying adjectives and adverbs (e.g. “most,” “really,” “essential,” “important,” “fascinating”).

    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
    taya
    -1.73
    entown
    -1.70
    uded
    -1.69
    aneity
    -1.62
    olving
    -1.61
    tya
    -1.59
    usions
    -1.56
    yship
    -1.52
    hever
    -1.52
    tych
    -1.51
    POSITIVE LOGITS
     thing
    2.34
     is
    2.25
    '
    1.93
     was
    1.82
     time
    1.78
    "
    1.70
     way
    1.67
     are
    1.67
     part
    1.55
     section
    1.52
    Activations Density 0.058%

    No Known Activations