© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 124017
    Prev
    Next
    INDEX
    Explanations

    numbers and roman numerals

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

    The neuron selectively activates on numeric tokens that include a decimal point (i.e. floating-point 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_10/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
    ℣
    -1.43
     behulp
    -1.42
    ଢ
    -1.38
     persoonlijke
    -1.38
    лимпий
    -1.31
     spelers
    -1.30
    醅
    -1.30
     BIBLIO
    -1.30
    oglobin
    -1.29
     másik
    -1.29
    POSITIVE LOGITS
     are
    1.69
     how
    1.43
    Доброго
    1.42
    並不是
    1.30
    已经
    1.28
     Это
    1.27
     Wochenende
    1.27
     것은
    1.27
     validés
    1.27
     Because
    1.27
    Activations Density 0.016%

    No Known Activations