© 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. 33067
    Prev
    Next
    INDEX
    Explanations

    location prepositions

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

    The neuron activates on nouns naming structural parts or anatomical components (e.g. projections, surfaces, muscles, barrels, stocks, wing veins, gills, walls, catheters, 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_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
    hadur
    -0.93
     arquitetura
    -0.92
     Dalam
    -0.91
    Dalam
    -0.90
    了
    -0.88
     während
    -0.87
    culosis
    -0.85
    INTAIN
    -0.84
    annt
    -0.83
    gether
    -0.82
    POSITIVE LOGITS
     near
    4.31
    near
    2.98
     between
    2.83
     at
    2.81
     below
    2.61
     above
    2.53
     just
    2.45
     Near
    2.38
    Near
    2.33
     près
    2.14
    Activations Density 0.179%

    No Known Activations