© 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-3-12B
    3. 24-GEMMASCOPE-2-RES-16K
    4. 12919
    Prev
    Next
    INDEX
    Explanations

    chapter names

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

    The neuron strongly activates on numerical tokens—especially specific figures, years, counts, and percentages.

    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
    getCql
    0.75
     Orden
    0.72
     swojego
    0.72
     Klicken
    0.70
    ای
    0.68
     Конечно
    0.68
     unittest
    0.66
    cuenta
    0.65
    ✙
    0.65
     flujo
    0.65
    POSITIVE LOGITS
    ھی
    0.78
     côt
    0.76
    iawan
    0.75
    inerary
    0.71
    nden
    0.70
    нын
    0.70
    anç
    0.69
    ilda
    0.69
    uba
    0.69
    ne
    0.68
    Activations Density 0.000%

    No Known Activations