© 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. 14070
    Prev
    Next
    INDEX
    Explanations

    Component, deletion, dismiss

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

    The neuron consistently fires on longer technical action–oriented words (especially code/API keywords) that carry common English verb‐to‐noun or adjective suffixes (–ion, –ment, –able, –ize, 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
    viamo
    -1.63
     "--
    -1.54
    举动
    -1.52
    ㄸ
    -1.49
     sufre
    -1.48
     exatamente
    -1.45
     únicamente
    -1.41
    ."
    -1.41
     diretamente
    -1.38
     hetzelfde
    -1.38
    POSITIVE LOGITS
     остальные
    1.70
    a
    1.58
     schönen
    1.54
    your
    1.48
     inspiradoras
    1.45
    m
    1.40
     fantástica
    1.38
    噚
    1.38
     normales
    1.38
     divertidas
    1.36
    Activations Density 0.000%

    No Known Activations