© 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. 22-GEMMASCOPE-RES-131K
    7. 128620
    Prev
    Next
    INDEX
    Explanations

    The neuron preferentially activates on longer, multi-syllabic content words (i.e. fairly long or formal terms) rather than short function words.

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

    useful to; cases where; remark leads; assumption that

    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
     the
    -1.16
     تأثير
    -0.91
    とおり
    -0.90
    公民
    -0.89
    ようだ
    -0.88
     اللاعب
    -0.87
     RadioButton
    -0.87
     капю
    -0.86
     =
    -0.85
     Parfum
    -0.84
    POSITIVE LOGITS
     všet
    1.16
     budeme
    1.14
     máme
    1.12
    ════
    1.07
     nejsou
    1.05
     budete
    1.05
    =[];
    1.05
     rů
    1.03
     sogar
    1.01
    nischer
    1.01
    Activations Density 0.004%

    No Known Activations