© 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. 12518
    Prev
    Next
    INDEX
    Explanations

    explaining anxieties, categories, functions, diseases

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

    The neuron fires strongly on hyphens used as title/list-item separators (i.e. the “–” dash marking off subtitles or appended clauses).

    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
    غ
    0.86
    น
    0.81
    َا
    0.78
    па
    0.77
     consecuencia
    0.77
    х
    0.76
    ע
    0.76
     regarded
    0.75
    小学校
    0.75
    POSITIVE LOGITS
     Minn
    0.83
    iation
    0.77
    raint
    0.77
    Repository
    0.76
    ycling
    0.74
    ällt
    0.74
    roup
    0.73
    atile
    0.71
    ått
    0.71
    orney
    0.71
    Activations Density 0.001%

    No Known Activations