© 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. 31-GEMMASCOPE-2-RES-16K
    7. 13208
    Prev
    Next
    INDEX
    Explanations

    i understand you might be exploring

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

    This neuron lights up on self-referential “I know” or “I can see”-style phrases where the author states their own knowledge or belief.

    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
    是否
    0.76
    是否存在
    0.71
     whether
    0.68
    whether
    0.68
    真相
    0.64
    是否有
    0.61
    facts
    0.60
     apakah
    0.60
    UNKNOWN
    0.60
    ABOUT
    0.59
    POSITIVE LOGITS
     Eindruck
    0.77
     फ्रॉम
    0.72
     frustration
    0.71
     reluctant
    0.70
     جذاب
    0.70
     gevoel
    0.69
     feeling
    0.67
     sentir
    0.66
     Interesse
    0.66
     preocupación
    0.66
    Activations Density 0.094%

    No Known Activations