© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-3-12B
    3. 24-GEMMASCOPE-2-RES-16K
    4. 3236
    Prev
    Next
    INDEX
    Explanations

    The neuron fires strongly on two‐word noun–noun compounds that denote a distinct section, area, unit, or subgroup (e.g. “garden areas,” “breakfast room,” “display units,” “staff members,” “articulation point”).

    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
    という
    0.55
    溼
    0.54
    僖
    0.54
     संस्करण
    0.53
    ඤ
    0.53
    Puedes
    0.52
     みたい
    0.52
    ISSION
    0.52
    وأ
    0.52
     グ
    0.51
    POSITIVE LOGITS
     itself
    0.63
    вался
    0.62
    承担
    0.61
     vary
    0.60
     अत्य
    0.60
     varies
    0.58
     fluctuate
    0.57
    दारों
    0.56
    强大
    0.56
     comprise
    0.55
    Activations Density 0.000%

    No Known Activations