© 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. 34-GEMMASCOPE-RES-131K
    4. 107246
    Prev
    Next
    INDEX
    Explanations

    Pokémon game titles

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

    The neuron strongly activates on occurrences of “Pokémon” (and related franchise/game titles like “Pokémon Go,” “Pokémon Black and White,” “Poké Ball,” etc.), i.e. mentions of the Pokémon brand.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
    fadeIn
    -0.76
    arities
    -0.76
     Anambra
    -0.73
    Chen
    -0.70
    횡
    -0.69
    ║
    -0.69
     getOrder
    -0.67
    dff
    -0.67
     schaffen
    -0.67
    チキン
    -0.67
    POSITIVE LOGITS
     Sword
    1.37
    Sword
    1.29
     Dyna
    1.20
     dyna
    1.17
    Dyna
    1.08
    sword
    0.95
     Gig
    0.94
     Shield
    0.92
    Scarlet
    0.88
     Schwä
    0.87
    Activations Density 0.010%

    No Known Activations