© 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
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 12951
    Prev
    Next
    INDEX
    Explanations

    type classification

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

    The neuron fires on the names of items when they’re being listed or enumerated (e.g. “chamber,” “pots,” “urinals,” “sandal,” “buckle,” “polyamory,” “mortar,” “grout,” “pouches,” “Citrus,” “Orange,” “Lemon,” etc.).

    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
     ngo
    -0.99
     ajuta
    -0.98
     alojamiento
    -0.96
    ův
    -0.96
     ubicación
    -0.95
     amazing
    -0.95
    owane
    -0.93
    ArgsConstructor
    -0.93
    veien
    -0.92
     COMPUT
    -0.92
    POSITIVE LOGITS
     or
    1.61
    Matériau
    1.14
     is
    1.14
    Our
    1.09
     type
    1.05
    Both
    1.01
     When
    0.98
    комнат
    0.94
     slight
    0.94
     Those
    0.94
    Activations Density 0.153%

    No Known Activations