© 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. 13644
    Prev
    Next
    INDEX
    Explanations

    loyal

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

    The neuron strongly activates on uncommon capitalized tokens—especially proper names of people, places, or organizations.

    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
     Vocational
    0.61
    trivia
    0.56
    Voc
    0.54
    बिट
    0.54
    र्नर
    0.52
    声
    0.52
    秘
    0.52
    讷
    0.52
    聲
    0.52
     Voc
    0.51
    POSITIVE LOGITS
     ло
    0.78
     Ло
    0.71
     Lob
    0.67
    Ло
    0.66
     Lola
    0.65
     Lo
    0.63
     allo
    0.63
    лло
    0.62
     lo
    0.60
     Lom
    0.60
    Activations Density 0.333%

    No Known Activations