© 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. 22-GEMMASCOPE-RES-131K
    7. 130959
    Prev
    Next
    INDEX
    Explanations

    The neuron fires on tokens naming validation or schema‐constraint directives (e.g. NotEmpty, required, presence, minItems/minimum/max, data-parsley-*, etc.).

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

    validation patterns and messages

    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
     these
    -1.16
     such
    -0.94
    ляр
    -0.93
     even
    -0.90
     nawet
    -0.90
     which
    -0.90
     inaweza
    -0.88
     how
    -0.88
     that
    -0.87
    而
    -0.85
    POSITIVE LOGITS
     salonu
    1.25
    śród
    1.20
    Gön
    1.18
     personajes
    1.18
     celebrid
    1.11
    imply
    1.09
     promesa
    1.06
    Asociación
    1.05
    plona
    1.04
     locura
    1.04
    Activations Density 0.003%

    No Known Activations