© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 13893
    Prev
    Next
    INDEX
    Explanations

    activation key

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

    This neuron detects modal and auxiliary verbs (e.g. contractions like ’re, ’ll, and words like “able,” “would,” “could,” “like” when used to express possibility or future action).

    oai_token-act-pair · o4-mini
    Triggered 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
    Negative Logits
     азо
    0.66
     মনোযোগ
    0.63
    磴
    0.61
    vVertex
    0.61
     estadio
    0.61
     وی
    0.60
     میلی
    0.59
     desempeño
    0.59
    迠
    0.58
    chyné
    0.57
    POSITIVE LOGITS
     simply
    0.64
     world
    0.58
     remember
    0.55
     Sunday
    0.54
     legal
    0.54
    Sun
    0.53
     worlds
    0.53
     unseen
    0.53
     dünyaya
    0.53
     simplement
    0.52
    Activations Density 0.001%

    No Known Activations