© 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. 65214
    Prev
    Next
    INDEX
    Explanations

    structural properties and data fields

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

    This neuron identifies JSON (or similar) object field names—i.e., keys before the colon in key–value data structures.

    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
     and
    -1.01
    :>
    -0.88
     meisten
    -0.87
     eerst
    -0.86
     aynı
    -0.85
    softc
    -0.85
     workaround
    -0.83
    Boletín
    -0.82
    ższy
    -0.79
     monstros
    -0.79
    POSITIVE LOGITS
     disponibilités
    1.12
     consisting
    0.99
     целом
    0.98
     including
    0.93
    комплек
    0.93
    📐
    0.91
     trio
    0.90
     includes
    0.90
     exclusivos
    0.89
     confortável
    0.89
    Activations Density 0.067%

    No Known Activations