© 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. 128783
    Prev
    Next
    INDEX
    Explanations

    The neuron strongly activates on label‐like tokens in code or markup—i.e. element/tag names, attribute names, and object‐property keys.

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

    attribute assignments

    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
     Popis
    -1.19
     this
    -1.18
     popis
    -1.14
    Nachteile
    -1.12
     both
    -1.12
    鯔
    -1.09
     bicicletas
    -1.09
    Odpowiedz
    -1.08
     Usually
    -1.07
     biela
    -1.07
    POSITIVE LOGITS
     also
    1.52
    r
    1.38
     durante
    1.30
    is
    1.26
    o
    1.25
    lo
    1.23
    one
    1.22
    ch
    1.22
    de
    1.21
    p
    1.20
    Activations Density 0.094%

    No Known Activations