© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 71878
    Prev
    Next
    INDEX
    Explanations

    I wanted

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

    The neuron strongly responds to numeric tokens—dates, times, event numbers, measurements—i.e. any instance of 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_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
     ihrer
    -1.42
     their
    -1.35
    เค้า
    -1.32
     are
    -1.23
     их
    -1.20
     ihren
    -1.20
     они
    -1.18
     mereka
    -1.18
    amkeit
    -1.18
    目前
    -1.17
    POSITIVE LOGITS
     my
    2.25
     our
    1.88
     I
    1.69
     моих
    1.32
     tensione
    1.31
     traseira
    1.30
     mojej
    1.28
     моей
    1.27
     me
    1.24
     escaleras
    1.23
    Activations Density 0.009%

    No Known Activations