© 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. 8795
    Prev
    Next
    INDEX
    Explanations

    enemy of

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

    The neuron is a paragraph/segment‐boundary detector—it fires strongly at tokens that mark the start (or end, like “?”) of a new text segment.

    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
    Missense
    0.81
    Recording
    0.78
     సాహిత
    0.78
     Steffen
    0.78
    م
    0.77
    Με
    0.75
     القلب
    0.75
    onnements
    0.75
     couverts
    0.74
    Actress
    0.73
    POSITIVE LOGITS
     held
    0.77
     violet
    0.73
     bush
    0.73
     yellow
    0.72
     mite
    0.72
    stag
    0.72
     victoria
    0.72
     sería
    0.71
     lug
    0.70
    /
    0.70
    Activations Density 0.001%

    No Known Activations