© 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. 10-GEMMASCOPE-RES-131K
    7. 71470
    Prev
    Next
    INDEX
    Explanations

    silverlight

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

    This neuron detects occurrences of the string “light,” especially in compound names or titles (i.e. words ending in “light”).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
    }$
    -2.91
     are
    -2.36
    BURGH
    -1.96
     inadvertently
    -1.88
    江西
    -1.88
    虚空
    -1.86
     It
    -1.84
    cat
    -1.82
    شارك
    -1.81
     Certainly
    -1.77
    POSITIVE LOGITS
    已经
    2.64
    跸
    2.30
    light
    2.28
    ers
    2.22
    くらいの
    2.19
    
    2.17
     mencionados
    2.11
     vistos
    2.05
     antidesliz
    2.05
     trabaja
    2.02
    Activations Density 0.147%

    No Known Activations