© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 27883
    Prev
    Next
    INDEX
    Explanations

    specific numbers and letters

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

    The neuron strongly responds to rare or domain-specific content words (e.g. low-frequency technical or specialised nouns).

    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
    imbo
    -0.96
     `
    -0.84
     [
    -0.82
     our
    -0.78
     both
    -0.76
    ByPrimaryKey
    -0.75
    e
    -0.73
     but
    -0.72
     apiKey
    -0.72
    o
    -0.72
    POSITIVE LOGITS
     هاي
    0.97
     describir
    0.90
    Tunes
    0.90
     transgression
    0.88
     satunya
    0.85
    Umar
    0.85
    angga
    0.84
    】：
    0.84
    ifikationen
    0.83
    সূত্র
    0.82
    Activations Density 0.341%

    No Known Activations