© 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. 12244
    Prev
    Next
    INDEX
    Explanations

    image source settings

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

    This neuron chiefly detects occurrences of the word “settings” (as in menu or configuration instructions).

    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
     тобі
    0.57
    同士
    0.52
    ৬০
    0.52
     pieces
    0.52
     natomiast
    0.52
    ‟
    0.52
    маш
    0.51
     واخ
    0.50
    麿
    0.50
     here
    0.49
    POSITIVE LOGITS
    ̚
    0.62
     getVideo
    0.60
     paquete
    0.59
    举报
    0.59
     微
    0.58
     富
    0.56
    ्ज
    0.55
     Pra
    0.54
     PRA
    0.54
    ൽ
    0.54
    Activations Density 0.000%

    No Known Activations