© 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. 95827
    Prev
    Next
    INDEX
    Explanations

    proper names and titles

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

    The neuron primarily activates on numeric tokens and measurements—years, stats, quantities, percentages, and decimal values.

    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
     in
    -0.94
    ktes
    -0.82
     travaillons
    -0.78
    戛
    -0.78
    ############
    -0.75
     so
    -0.75
     when
    -0.75
     produisons
    -0.74
     before
    -0.73
    特的
    -0.73
    POSITIVE LOGITS
     capaces
    0.94
     heartwarming
    0.90
    tión
    0.90
    うわ
    0.90
     leſs
    0.89
     دلم
    0.88
    zepte
    0.87
     hoang
    0.87
     substantive
    0.86
     твой
    0.85
    Activations Density 0.113%

    No Known Activations