© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 49544
    Prev
    Next
    INDEX
    Explanations

    entering or being in town

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

    The neuron fires on tokens that mark movement or entering a place—especially prepositions like “into,” “in,” or “back” when used in travel/arrival contexts.

    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
    現在は
    -0.79
    ショナル
    -0.72
     cylinder
    -0.72
    |'.'
    -0.71
    què
    -0.71
     成功
    -0.71
     a
    -0.69
    portion
    -0.69
     работы
    -0.68
     роботи
    -0.68
    POSITIVE LOGITS
     town
    5.81
    town
    3.11
     TOWN
    2.98
    Town
    2.64
     Town
    2.56
    TOWN
    2.52
     towns
    2.19
    町
    2.16
     city
    2.13
     городе
    1.95
    Activations Density 0.038%

    No Known Activations