© 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. 31-GEMMASCOPE-2-RES-16K
    7. 10790
    Prev
    Next
    INDEX
    Explanations

    instructions, or action steps

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

    The neuron detects standalone, capitalized label or header tokens that introduce document sections (e.g. “Product,” “Address,” “Message,” “Features,” 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_31_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
    묽
    0.59
    アクセサリー
    0.57
    startButton
    0.55
    淢
    0.54
    BeginInit
    0.54
    考生
    0.53
    мышлен
    0.52
    িমুখে
    0.52
    ༸
    0.52
     locom
    0.52
    POSITIVE LOGITS
     Ashley
    0.64
    Cited
    0.61
     Cited
    0.59
     Fred
    0.58
    ங்கும்
    0.57
    ify
    0.57
     Flip
    0.55
     Scott
    0.55
     Stuart
    0.55
    Edit
    0.54
    Activations Density 0.000%

    No Known Activations