© 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. 85142
    Prev
    Next
    INDEX
    Explanations

    cars and vehicles

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

    The neuron activates on numeric tokens—especially floating-point numbers or measurements (e.g. “19.25,” “28.5,” “52.75,” etc.).

    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
    bringen
    -1.66
    炘
    -1.64
    richtung
    -1.63
    熺
    -1.59
    嫆
    -1.54
    namely
    -1.52
    selben
    -1.51
     écout
    -1.50
    anches
    -1.49
    iettes
    -1.48
    POSITIVE LOGITS
     &
    2.08
    >
    1.62
     rita
    1.59
    ’
    1.59
     iveco
    1.57
     For
    1.54
    :
    1.54
     elaborated
    1.52
    方法
    1.45
    );
    1.41
    Activations Density 0.006%

    No Known Activations