© 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. 1505
    Prev
    Next
    INDEX
    Explanations

    The neuron fires most strongly on tokens containing digits (numbers, version/file identifiers, dates, decimal values, etc.).

    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
     
    1.27
     Ayrıca
    1.27
     Özellikle
    1.23
     えっと
    1.21
     
    1.20
     Infatti
    1.18
     Júnior
    1.18
     Außerdem
    1.17
     Porém
    1.16
     Çünkü
    1.16
    POSITIVE LOGITS
    _
    2.52
    -
    1.99
    \_
    1.38
    ・
    1.32
    __
    1.30
    -(
    1.13
    -{
    1.09
    _%
    1.06
    get
    1.03
    point
    1.03
    Activations Density 0.168%

    No Known Activations