© 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. 13320
    Prev
    Next
    INDEX
    Explanations

    looping or increasing counts

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

    Explanation of neuron 4 behavior: the main thing this neuron does is find numeric expressions and quantitative/statistical data in the text.

    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
    );//
    0.68
    []){
    0.66
    等于
    0.66
    []=
    0.64
    😖
    0.63
    ;//
    0.62
    ];//
    0.61
    )//
    0.60
     Swal
    0.59
    =(
    0.58
    POSITIVE LOGITS
     dear
    0.59
     dearest
    0.59
     Please
    0.57
     do
    0.55
     Do
    0.54
    nPlease
    0.54
     pü
    0.54
     pö
    0.53
     escalating
    0.52
    Please
    0.52
    Activations Density 0.000%

    No Known Activations