© 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. 5091
    Prev
    Next
    INDEX
    Explanations

    breakdown

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

    The neuron responds to dash (–/‐) punctuation tokens, especially when they occur as separators before numbers or in dateline‐style 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
    ні
    0.86
     flavon
    0.83
     disson
    0.79
    tariff
    0.75
    e
    0.75
     idade
    0.75
    tarif
    0.73
    ной
    0.71
     pept
    0.71
     atheros
    0.71
    POSITIVE LOGITS
     друзья
    0.70
    WEBPACK
    0.69
    ＂
    0.68
     Gino
    0.68
    ん
    0.68
     unterwegs
    0.66
    據
    0.66
    gevens
    0.66
    ㄨ
    0.66
    কলকাতা
    0.66
    Activations Density 0.000%

    No Known Activations