© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 32872
    Prev
    Next
    INDEX
    Explanations

    also too likewise

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

    This neuron activates primarily on numeric tokens and quantitative measurements (e.g. durations, counts, and values in scientific or listing contexts).

    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
     čer
    -0.80
     Auftritt
    -0.80
    devamını
    -0.77
    ऍ
    -0.75
    还是
    -0.75
     مالی
    -0.75
    Rather
    -0.74
    Australie
    -0.73
     asociación
    -0.73
    還是
    -0.72
    POSITIVE LOGITS
     ebenfalls
    1.49
     also
    1.33
     тоже
    1.32
     نیز
    1.25
     również
    1.16
     likewise
    1.01
     too
    0.98
     també
    0.93
     também
    0.93
     также
    0.92
    Activations Density 0.206%

    No Known Activations