© 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. 128081
    Prev
    Next
    INDEX
    Explanations

    The neuron flags tokens used as keys or property names in structured data (e.g. JSON object keys, JavaScript/HTML template properties, config/map keys).

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

    data keys

    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
     him
    -0.98
    修士
    -0.97
    从
    -0.96
     his
    -0.93
    可以
    -0.93
     can
    -0.93
     and
    -0.90
    psie
    -0.89
    duate
    -0.87
    将
    -0.86
    POSITIVE LOGITS
    特别的
    1.26
    接下来的
    1.20
    uló
    1.16
    比べ
    1.13
    Recept
    1.05
    Lá
    1.05
     personnaliser
    1.04
     Inscrivez
    1.02
    Прият
    1.01
    Кстати
    1.01
    Activations Density 0.008%

    No Known Activations