© 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. 14029
    Prev
    Next
    INDEX
    Explanations

    questions about questions

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

    The neuron fires most strongly on the very first meaningful word of each new text block or paragraph, effectively marking segment or paragraph boundaries.

    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
    ت
    1.00
    ل
    0.80
    ת
    0.76
    ה
    0.76
    Ptr
    0.75
    _
    0.75
    るので
    0.73
    ща
    0.71
    än
    0.71
    =[]
    0.71
    POSITIVE LOGITS
     outbursts
    1.03
     médian
    0.96
     dosages
    0.88
     Garantie
    0.86
     tailles
    0.86
     enquête
    0.83
     doses
    0.81
     occupiers
    0.81
     orale
    0.81
     outburst
    0.81
    Activations Density 0.001%

    No Known Activations