© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 111049
    Prev
    Next
    INDEX
    Explanations

    Superlative adjectives

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

    This neuron fires on big, flashy promotional adjectives (like “Awesome,” “Amazing,” “Fabulous,” “Wonderful,” etc.) typically used in titles or hype language.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
     Nunca
    -0.93
    WHILE
    -0.87
    ISHMENT
    -0.85
     WHILE
    -0.84
     dente
    -0.84
     laboratoire
    -0.84
     Diret
    -0.81
    ardrop
    -0.80
    所有的
    -0.78
     ilyen
    -0.77
    POSITIVE LOGITS
     Amazing
    1.07
     as
    1.02
     Fantastic
    1.00
     indeed
    0.96
     resultados
    0.95
    Indeed
    0.94
    哉
    0.93
     Awesome
    0.91
     Wonderful
    0.91
     pretty
    0.90
    Activations Density 0.037%

    No Known Activations