© 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. 13466
    Prev
    Next
    INDEX
    Explanations

    mathematical expressions with variables

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

    The neuron strongly lights up on uppercase single‐letter symbols and matrix expressions in mathematical formulas, i.e. when the text is describing matrices, identity matrices, or matrix multiplications.

    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
     ∈
    0.79
     ÷
    0.75
    ∝
    0.74
    ∈
    0.73
    ÷
    0.70
    ∩
    0.67
    =(
    0.66
     ×
    0.65
    =(-
    0.62
     =
    0.61
    POSITIVE LOGITS
    狈
    0.63
     সবাইকে
    0.62
     публику
    0.59
    他就
    0.57
     najwięks
    0.56
     publice
    0.55
     வழங்கு
    0.55
    öffent
    0.54
     લોકોને
    0.54
    ছেলে
    0.54
    Activations Density 0.004%

    No Known Activations