© 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. 129182
    Prev
    Next
    INDEX
    Explanations

    The neuron fires on programming identifier tokens (names of variables, functions, constants, etc.) rather than ordinary words.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    type casts, function types, number defaults

    np_acts-logits-general · gemini-2.5-flash-lite
    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
     there
    -1.16
     There
    -1.00
    的一些
    -0.94
     Therefore
    -0.93
     总
    -0.93
    belongsTo
    -0.92
    mustang
    -0.92
     ehe
    -0.91
     hierdie
    -0.91
     Depending
    -0.88
    POSITIVE LOGITS
    且
    1.34
     lyck
    1.13
     zwłaszcza
    1.09
    ISupport
    1.08
    ,‎
    1.07
    cinating
    1.05
     و
    1.02
    และ
    1.01
     hlavně
    0.99
     sauf
    0.98
    Activations Density 0.041%

    No Known Activations