© 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. 33026
    Prev
    Next
    INDEX
    Explanations

    test, assert, expect

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

    This neuron flags programming test and assertion keywords (e.g. “expect,” “assert,” “test,” constructor/cause names) in code.

    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
    支部
    -1.02
    ADUATE
    -0.98
    ciona
    -0.98
     offiziell
    -0.98
    aneity
    -0.93
    acionais
    -0.90
     dimaksud
    -0.90
     Pleistocene
    -0.90
    lijkheid
    -0.90
     a
    -0.90
    POSITIVE LOGITS
     descobrir
    1.02
     reduzir
    0.98
     começar
    0.98
     ocorrer
    0.94
     então
    0.90
     használat
    0.90
    diteur
    0.88
    マス
    0.88
     söyledi
    0.88
    也没有
    0.87
    Activations Density 0.004%

    No Known Activations