© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 2908
    Prev
    Next
    INDEX
    Explanations

    The neuron fires most strongly on concrete object nouns (especially tools or instrument names).

    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
    و
    0.75
    further
    0.69
    त्‍
    0.66
     पेंशन
    0.63
     क्रिप्टोकर
    0.63
    শরণ
    0.63
    on
    0.62
    ప్ర
    0.62
    وون
    0.61
    =\"
    0.61
    POSITIVE LOGITS
     podía
    0.78
     kost
    0.77
    ものが
    0.75
     programación
    0.75
     mouthful
    0.74
     comenzar
    0.74
     antiguo
    0.74
     supuesto
    0.73
     Spectral
    0.73
     chalet
    0.73
    Activations Density 0.000%

    No Known Activations