© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 93419
    Prev
    Next
    INDEX
    Explanations

    programming constructs

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

    The neuron fires on loop‐header syntax in code (e.g. the “while(…)” or “for(…)” constructs), especially on the closing parenthesis of those loop conditions.

    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
     houden
    -0.97
    沒有
    -0.93
     keberadaan
    -0.92
     available
    -0.89
     who
    -0.87
     a
    -0.85
    谡
    -0.85
     kehadiran
    -0.84
    ravel
    -0.83
    Places
    -0.82
    POSITIVE LOGITS
    Rivers
    0.81
    greenrobot
    0.80
     clock
    0.80
     mě
    0.80
    :”
    0.77
    anin
    0.77
    }
    0.76
    дям
    0.72
     Cili
    0.72
     مع
    0.71
    Activations Density 0.001%

    No Known Activations