© 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-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 5351
    Prev
    Next
    INDEX
    Explanations

    shapes and structures

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

    The neuron fires strongly on proper names and titles—brand names, product names, organizations, and other named entities.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_24_width_16k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     menilai
    0.88
     keselamatan
    0.81
    賄
    0.80
     valutazione
    0.79
     качества
    0.79
     setempat
    0.78
    atthaya
    0.78
    贖
    0.76
    価値
    0.75
    azione
    0.74
    POSITIVE LOGITS
     rectangles
    1.52
     rectangular
    1.44
     polygons
    1.35
     diameters
    1.34
     horizontally
    1.34
     hexagons
    1.32
     elongated
    1.31
     ellipses
    1.30
     girder
    1.27
     rectangle
    1.27
    Activations Density 12.311%

    No Known Activations