© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 129927
    Prev
    Next
    INDEX
    Explanations

    This neuron responds to programming‐style tokens—e.g. CSS hex color codes, punctuation like semicolons, and code identifiers (capitalized or symbol-rich tokens).

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

    project summary, studies, questions

    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
    tellar
    -1.20
    actéris
    -1.18
    onomie
    -1.15
    mpton
    -1.14
    tembro
    -1.13
    يكب
    -1.09
     useAppContext
    -1.07
     Descripción
    -1.05
     Método
    -1.05
    [:,:,
    -1.05
    POSITIVE LOGITS
     такая
    0.96
     solcher
    0.95
     atento
    0.93
     Gre
    0.91
     Revenue
    0.90
    T
    0.89
    ഊ
    0.88
    4
    0.88
     usual
    0.87
    3
    0.85
    Activations Density 0.000%

    No Known Activations