© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 14046
    Prev
    Next
    INDEX
    Explanations

    the number 3

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

    The neuron fires almost exclusively on the single digit “3,” i.e. it detects occurrences of the numeral 3 in the text.

    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
    replyreport
    -2.94
    dersfield
    -2.45
     participó
    -2.42
    т
    -2.42
    髣
    -2.39
    after
    -2.38
    irical
    -2.33
     regresó
    -2.30
    不论
    -2.27
     trabajó
    -2.25
    POSITIVE LOGITS
    邶
    2.78
    ୮
    2.67
    ỽ
    2.59
    硖
    2.53
    棡
    2.52
    豳
    2.50
    ol
    2.34
    鼍
    2.23
     Our
    2.22
    鮰
    2.22
    Activations Density 0.034%

    No Known Activations