© 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. 12368
    Prev
    Next
    INDEX
    Explanations

    Urgent/Important and Communication

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

    The neuron is a bibliographic‐citation detector: it lights up on tokens that occur in formal reference entries (e.g. journal abbreviations, citation punctuation, and related metadata) rather than in running narrative.

    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
     операций
    0.77
    най
    0.75
     иных
    0.75
     цепо
    0.74
     другой
    0.71
    이었
    0.70
    dropdown
    0.69
    籤
    0.69
    média
    0.68
    ೇವೆ
    0.66
    POSITIVE LOGITS
    ل
    0.91
     जहा
    0.85
     Sebab
    0.84
    ورت
    0.83
    itung
    0.80
    it
    0.77
     ihm
    0.77
    istung
    0.77
     messen
    0.77
    as
    0.76
    Activations Density 0.000%

    No Known Activations