© 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-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 65219
    Prev
    Next
    INDEX
    Explanations

    left margin and padding

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

    The neuron activates on numeric literals (e.g. pixel, dp, or other measurement values) in style or layout declarations.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     we
    -1.20
    ↵↵↵
    -1.09
     should
    -1.09
     if
    -1.02
     墙
    -1.00
    Located
    -0.96
     [
    -0.96
    ,
    -0.96
     for
    -0.94
     questions
    -0.94
    POSITIVE LOGITS
     remporté
    1.16
     sorprend
    1.13
     accordé
    1.09
    绝望
    1.09
     consacré
    1.08
    
    1.07
     veranderen
    1.07
     nadru
    1.06
     tuvimos
    1.05
    gnac
    1.05
    Activations Density 0.003%

    No Known Activations