© 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. 16444
    Prev
    Next
    INDEX
    Explanations

    code and technical terms

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

    The neuron activates on numeric tokens—especially floating-point numbers and other digit strings (e.g., “106.5,” “97.5,” “80.5”).

    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
    ització
    -1.00
     darbu
    -0.99
    inkite
    -0.93
    <()>
    -0.92
    ֹת
    -0.90
    Prof
    -0.88
     of
    -0.87
    Ẽ
    -0.86
    AxisAlignment
    -0.85
     kjent
    -0.81
    POSITIVE LOGITS
    meteries
    0.98
    Осно
    0.93
    しかも
    0.86
    or
    0.82
    تيجة
    0.82
     encontrar
    0.80
    and
    0.80
    versus
    0.79
    jTable
    0.79
     inserir
    0.76
    Activations Density 0.056%

    No Known Activations