© 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. 26653
    Prev
    Next
    INDEX
    Explanations

    parsers and data structures

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

    The neuron acts as a detector for code‐style tokens—variable names, acronyms, JSON/XML identifiers, CamelCase or all-caps programming terms—within mixed text.

    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
    :
    -1.15
     only
    -1.05
     can
    -1.04
     some
    -0.96
    ,
    -0.94
     Each
    -0.87
     ONLY
    -0.86
     on
    -0.86
    ;
    -0.85
     make
    -0.83
    POSITIVE LOGITS
    そういえば
    1.10
     菇
    1.08
     wyja
    0.99
    
    0.95
     njego
    0.95
     volcan
    0.93
    צב
    0.92
     jego
    0.92
     anuncia
    0.91
     forbids
    0.91
    Activations Density 0.151%

    No Known Activations