© 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. 34222
    Prev
    Next
    INDEX
    Explanations

    alternatives with types like array, string, integer

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

    The neuron fires on mentions of programming data types, classes, or other type names (i.e. tokens that name types or classes).

    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
    ėje
    -1.00
    ÂN
    -0.96
    AppState
    -0.94
    ksessa
    -0.93
    LogUtil
    -0.90
     dung
    -0.90
    EURO
    -0.88
     aliviar
    -0.88
     nă
    -0.87
    确认
    -0.86
    POSITIVE LOGITS
     or
    1.71
     anything
    1.38
    anything
    1.16
     cualquier
    1.14
    ilions
    1.09
    icace
    1.07
     또는
    1.07
    也可以
    1.07
     Alternatively
    1.06
    shmi
    1.06
    Activations Density 0.028%

    No Known Activations