© 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
    2. Gemma-3-12B
    3. 24-GEMMASCOPE-2-RES-16K
    4. 5945
    Prev
    Next
    INDEX
    Explanations

    general/purely/overly descriptive information

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

    The neuron fires on adverbs and adjectives used as discourse markers or intensifiers (e.g. “well,” “right here,” “simply,” “constantly,” “extra,” “more,” “quite,” etc.).

    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
     अधिकांश
    0.53
    ж
    0.53
     crocodiles
    0.50
    صبح
    0.50
    вере
    0.49
     undisputed
    0.49
    ται
    0.48
     inviol
    0.48
    Ʌ
    0.48
     quartet
    0.48
    POSITIVE LOGITS
     มัน
    0.61
     Huge
    0.61
    มัน
    0.58
     Suggestions
    0.57
    ,
    0.53
    并且
    0.53
     Worte
    0.53
    sma
    0.52
     Corporations
    0.52
    înes
    0.52
    Activations Density 0.261%

    No Known Activations