© 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. 72429
    Prev
    Next
    INDEX
    Explanations

    special

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

    The main thing this neuron does is detect occurrences of the adjective “special.”

    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
     lã
    -1.38
    yship
    -1.24
    慟
    -1.20
    睨
    -1.19
     bekämp
    -1.18
    莛
    -1.16
    𓁹
    -1.16
    зык
    -1.14
     assurent
    -1.13
    揶
    -1.13
    POSITIVE LOGITS
     special
    2.08
     designed
    1.53
    特別
    1.52
     specific
    1.49
     for
    1.46
     unique
    1.43
    special
    1.43
     khusus
    1.36
     специаль
    1.34
     особы
    1.34
    Activations Density 0.056%

    No Known Activations