© 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. 8571
    Prev
    Next
    INDEX
    Explanations

    foreign language characters

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

    This neuron fires most strongly on tokens at the very start of a document (with activation decaying as you move further into the text).

    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
    ちは
    0.80
    Result
    0.71
    _
    0.71
    🙉
    0.71
    _,
    0.68
    inar
    0.64
    _)
    0.64
    蒋
    0.64
    Principal
    0.64
    楽天
    0.64
    POSITIVE LOGITS
     svjet
    0.81
    ний
    0.79
    ემა
    0.77
     kuidas
    0.77
     tüket
    0.77
     crustaceans
    0.77
     tableware
    0.77
     skaters
    0.77
     thermocou
    0.76
     shrimp
    0.75
    Activations Density 0.000%

    No Known Activations