© 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. 1595
    Prev
    Next
    INDEX
    Explanations

    email drafts miss working

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

    The neuron strongly activates on low‐frequency or unusually formal tokens—i.e. words that are rare or specialized rather than common everyday vocabulary.

    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
     hva
    0.68
     (
    0.67
     What
    0.63
     what
    0.63
     lawsuit
    0.62
    http
    0.61
    —
    0.61
    What
    0.60
     hvad
    0.60
                 
    0.59
    POSITIVE LOGITS
     QuantityOnHand
    0.75
     おしゃれ
    0.75
     Disha
    0.74
     Анастасия
    0.72
     ServicePolicy
    0.71
    ungnya
    0.71
     Фургала
    0.71
     Bildern
    0.71
     Фурга
    0.70
    <unused672>
    0.70
    Activations Density 0.008%

    No Known Activations