© 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. 64865
    Prev
    Next
    INDEX
    Explanations

    code with functions and arguments

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

    The neuron is looking for code’s opening‐parenthesis tokens (i.e. “(” in function definitions, calls, lambdas, etc.).

    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
    ږ
    -0.96
    ⟣
    -0.95
     erken
    -0.94
     intelligente
    -0.93
    kring
    -0.93
     поэтому
    -0.93
    readFileSync
    -0.91
     ilustracji
    -0.90
     olika
    -0.89
     passende
    -0.89
    POSITIVE LOGITS
     meint
    0.90
     documents
    0.89
     faithfully
    0.88
    処分
    0.88
     году
    0.87
    ELIA
    0.87
     erreur
    0.86
     würde
    0.86
    を実施
    0.84
    ược
    0.83
    Activations Density 0.006%

    No Known Activations