© 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. 75201
    Prev
    Next
    INDEX
    Explanations

    structured format definitions

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

    The neuron fires on field or key names (labels) in structured data or code—i.e. tokens that serve as schema/property identifiers.

    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
     their
    -1.18
     Weiterbildung
    -1.12
     when
    -1.00
     Makes
    -0.96
     what
    -0.96
     now
    -0.95
    cję
    -0.93
     легко
    -0.92
     literally
    -0.91
     there
    -0.91
    POSITIVE LOGITS
    liath
    0.98
    rawy
    0.96
     ステンレス
    0.96
     tı
    0.95
     fz
    0.93
    jan
    0.92
    Caracteristici
    0.91
    –
    0.90
     TECHN
    0.89
    tearDown
    0.89
    Activations Density 0.001%

    No Known Activations