© 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. 12732
    Prev
    Next
    INDEX
    Explanations

    in some formats

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

    The neuron fires on the principal “topic” word at the end of each section or paragraph, effectively picking out that section’s key content word.

    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
    ς
    0.84
     Erectile
    0.80
     Feed
    0.77
    Biz
    0.75
     Chum
    0.72
     Leather
    0.71
    Feed
    0.70
     তারি
    0.70
     Ingeniería
    0.70
    Date
    0.69
    POSITIVE LOGITS
    o
    0.86
    빴
    0.78
     الاخير
    0.77
    ponen
    0.77
     epile
    0.76
    rete
    0.75
    新た
    0.75
    വസ്ഥ
    0.75
    bersome
    0.75
     invertebr
    0.74
    Activations Density 0.000%

    No Known Activations