© 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. 130746
    Prev
    Next
    INDEX
    Explanations

    This neuron responds to numeric tokens (e.g. dates, scores, ages, rankings) in the text.

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

    news article markers

    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
    obrá
    -0.93
    вушка
    -0.92
     charbon
    -0.89
     toekomst
    -0.86
    hrungen
    -0.86
    auces
    -0.86
     Coher
    -0.84
    fährt
    -0.82
     carreaux
    -0.81
     gramos
    -0.81
    POSITIVE LOGITS
     different
    1.17
     these
    1.07
     with
    1.03
    ulico
    1.01
     previous
    0.95
     what
    0.94
     how
    0.92
     similar
    0.91
     any
    0.90
     evitare
    0.90
    Activations Density 0.002%

    No Known Activations