© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 11366
    Prev
    Next
    INDEX
    Explanations

    kernel module or login

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

    The neuron chiefly fires on past‐tense verb tokens (e.g. “included,” “required,” “was,” “found”).

    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
    aver
    0.63
    и
    0.62
    -<
    0.54
    ingham
    0.54
    Matcher
    0.54
    -
    0.53
     floppy
    0.51
    matcher
    0.51
    Admin
    0.50
    ↵↵↵
    0.50
    POSITIVE LOGITS
     Maugin
    0.67
    ুনিক
    0.58
     Treatise
    0.58
    angsa
    0.55
     রিপোর্টে
    0.53
     vitth
    0.52
    orsun
    0.52
     دوبارہ
    0.51
    纍
    0.51
    odeal
    0.51
    Activations Density 0.000%

    No Known Activations