© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 14811
    Prev
    Next
    INDEX
    Explanations

    chair, virus, @param, message%, whether, antarctica

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

    This neuron lights up on salient content words (not function words), i.e. relatively uncommon nouns and verbs carrying the main semantic load of a sentence.

    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
    다면
    0.84
     swojego
    0.84
     знаходиться
    0.81
     teljes
    0.80
    yleg
    0.80
    ٢
    0.80
    elipe
    0.80
     pulau
    0.77
    ન્યા
    0.77
     України
    0.77
    POSITIVE LOGITS
     cravings
    0.73
     Gems
    0.69
     proté
    0.68
    挾
    0.67
     influential
    0.67
     concomit
    0.66
     modifié
    0.66
     virulent
    0.65
     accuracies
    0.65
     sensit
    0.64
    Activations Density 0.001%

    No Known Activations