© 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. 131015
    Prev
    Next
    INDEX
    Explanations

    The neuron selectively lights up on tokens that are numeric literals or number-like values (integers or decimals).

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

    code, legal, technical, and numerical data

    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
    Према
    -0.96
    Dacă
    -0.84
    Então
    -0.83
     in
    -0.83
    Entonces
    -0.82
    &
    -0.81
    Apenas
    -0.81
    Được
    -0.81
     and
    -0.80
    Él
    -0.80
    POSITIVE LOGITS
     ſeveral
    1.00
    ();*/
    0.94
    /}.
    0.92
     "),
    0.90
    つける
    0.90
     می‌توانید
    0.89
    سجيلات
    0.88
    ✌
    0.86
     カーディガン
    0.86
    UKU
    0.85
    Activations Density 0.116%

    No Known Activations