© 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. 312
    Prev
    Next
    INDEX
    Explanations

    specific phrases like "Power Shift" or "The Samsung"

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

    The neuron is detecting numeric tokens (individual digits or numbers).

    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
     their
    -1.12
     for
    -1.00
    Their
    -0.98
     if
    -0.97
     awakeFromNib
    -0.92
     escrito
    -0.91
     GetEnumerator
    -0.91
     first
    -0.90
    راك
    -0.89
     çünkü
    -0.88
    POSITIVE LOGITS
     endeav
    0.99
     contemporaneous
    0.97
     EXCELL
    0.97
     quintessential
    0.95
     unsurpassed
    0.94
    níku
    0.93
     همچ
    0.93
     enraged
    0.92
     dále
    0.92
     fervent
    0.91
    Activations Density 0.000%

    No Known Activations