© 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. 129893
    Prev
    Next
    INDEX
    Explanations

    The neuron fires on content‐heavy, specialized terms or technical jargon (e.g. domain-specific nouns).

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

    layer, motor, brave, RAC, Group, Swansea, counter, bowl, memory

    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
     (!__
    -1.20
     valamint
    -1.12
    urator
    -1.09
    tiness
    -1.07
    umpad
    -1.06
     _____
    -1.05
    strual
    -1.05
    strator
    -1.04
    taya
    -1.03
    manteau
    -1.03
    POSITIVE LOGITS
     (
    1.31
    したので
    1.28
    horabuena
    1.16
     hés
    1.16
    oquia
    1.14
     favoris
    1.12
     contactez
    1.10
     on
    1.09
     éter
    1.09
     nécess
    1.05
    Activations Density 0.102%

    No Known Activations