© 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
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 1075
    Prev
    Next
    INDEX
    Explanations

    The neuron fires on connective function words (e.g., conjunctions and prepositions) that signal clause transitions or elaborations.

    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
     tiež
    0.35
     redux
    0.35
    een
    0.34
    spirational
    0.34
     மற்றும்
    0.33
     eds
    0.33
    âteau
    0.33
    .*;
    0.33
    .].
    0.33
     рамках
    0.32
    POSITIVE LOGITS
    不會
    0.46
    ن
    0.37
     tidak
    0.36
    不超过
    0.36
     નહીં
    0.35
     Tidak
    0.35
    不允许
    0.35
    不满
    0.35
    不说
    0.34
    不见
    0.34
    Activations Density 0.000%

    No Known Activations