© 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. 33626
    Prev
    Next
    INDEX
    Explanations

    code and news headlines

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

    The neuron fires on mis-encoded or out-of-place special/Unicode character sequences (e.g. stray “Â,” odd quotation marks, entity fragments, 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_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
     ahşap
    -1.09
     this
    -1.08
     że
    -1.05
     two
    -0.99
     orilla
    -0.94
    ;
    -0.88
     on
    -0.88
    kład
    -0.88
     I
    -0.87
     ganska
    -0.87
    POSITIVE LOGITS
    куль
    1.13
    ડી
    1.09
    Efter
    1.05
    фай
    1.05
    entee
    1.03
    schung
    1.03
     Marijuana
    1.02
    intos
    1.02
    ACIONES
    1.01
    澳洲
    1.01
    Activations Density 0.005%

    No Known Activations