© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 32858
    Prev
    Next
    INDEX
    Explanations

    foreign language

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

    The neuron is tuned to single words that label prominent states, events, or conditions (e.g. “fire,” “silence,” “respond,” “exposed,” “control,” “test,” “notice,” “unstable,” “problem”).

    oai_token-act-pair · o4-mini
    Triggered by @jyhe0408
    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
    Negative Logits
    følge
    -0.90
     新宿
    -0.85
    Friend
    -0.85
    éia
    -0.83
    ئے
    -0.81
     السكان
    -0.80
    iligt
    -0.78
    itarios
    -0.78
     květ
    -0.78
    ेद
    -0.77
    POSITIVE LOGITS
     its
    1.30
     pregunt
    0.90
    its
    0.89
     Enfer
    0.85
    terne
    0.85
    它的
    0.85
    صاص
    0.84
     российского
    0.81
    它
    0.80
     dessinée
    0.80
    Activations Density 0.131%

    No Known Activations