© 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. 128449
    Prev
    Next
    INDEX
    Explanations

    This neuron responds to unusually long words, i.e. tokens of high character‐length.

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

    tragic outcomes and deaths

    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
     Traum
    -0.88
     preferências
    -0.81
    Null
    -0.81
    ányi
    -0.81
     upheaval
    -0.81
     oči
    -0.80
    ッピー
    -0.79
    يع
    -0.79
     había
    -0.78
    zitter
    -0.78
    POSITIVE LOGITS
     fatal
    1.48
     fatally
    1.29
     died
    1.15
    fatal
    1.14
    Fatal
    1.06
     deaths
    1.06
     tragically
    1.03
     fateful
    1.01
     Fatal
    0.97
     tragedy
    0.96
    Activations Density 0.101%

    No Known Activations