© 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. 80028
    Prev
    Next
    INDEX
    Explanations

    boolean checks for null or defined

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

    The neuron strongly activates on boolean‐logic operators and related punctuation (e.g. “||”, “&&”, “or”, commas separating logical subexpressions).

    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
     pourront
    -1.09
     cerveau
    -0.98
    щита
    -0.97
     также
    -0.96
     getroffen
    -0.95
     Heilige
    -0.95
    okkal
    -0.94
     couvrir
    -0.94
    wandel
    -0.94
     glisser
    -0.94
    POSITIVE LOGITS
     in
    1.07
     on
    1.03
     but
    1.01
    哪怕
    0.98
     tenía
    0.98
     r
    0.96
     >
    0.96
    !
    0.96
     any
    0.94
     p
    0.94
    Activations Density 0.003%

    No Known Activations