© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 11164
    Prev
    Next
    INDEX
    Explanations

    programming languages and world

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

    The neuron strongly activates on numeric measurements given as decimal numbers (i.e. tokens containing digits and a decimal point).

    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
     Calibration
    0.55
    iculaire
    0.55
     sebab
    0.54
     Cal
    0.53
     Essentially
    0.53
    ↵
    0.53
     dos
    0.51
     gly
    0.50
    lema
    0.50
     reddit
    0.50
    POSITIVE LOGITS
    第七
    0.68
    playable
    0.65
    第五
    0.64
    নাট
    0.61
    ৌজের
    0.60
     McGr
    0.57
    覀
    0.57
    第八
    0.56
    FFFFFF
    0.55
    第六
    0.54
    Activations Density 0.000%

    No Known Activations