© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 128364
    Prev
    Next
    INDEX
    Explanations

    The neuron activates on standalone numeric tokens (integers, decimals, and fractions).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    numbers following punctuation or words like `cv`

    np_acts-logits-general · gemini-2.5-flash-lite
    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
     if
    -0.95
     which
    -0.91
     while
    -0.89
    ၆
    -0.88
     our
    -0.88
     when
    -0.88
     only
    -0.88
    loja
    -0.88
    bruno
    -0.87
     tocado
    -0.87
    POSITIVE LOGITS
     molti
    0.99
    four
    0.93
    orn
    0.92
     stalwart
    0.92
    below
    0.91
    NOT
    0.91
    DIS
    0.90
    ndash
    0.88
     mediocre
    0.88
    RO
    0.88
    Activations Density 0.004%

    No Known Activations