© 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. 33938
    Prev
    Next
    INDEX
    Explanations

    special tokens followed by punctuation

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

    The neuron fires on isolated code option names or constants (e.g. “copy”, “readonly”, “SET”) used as keyword arguments or parameters 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
     volverá
    -1.16
     already
    -1.16
     otorga
    -1.09
    makeat
    -1.05
     satria
    -1.02
     apresentam
    -1.02
     面包
    -1.01
     tokopedia
    -1.01
     メッセージ
    -1.00
    curate
    -0.99
    POSITIVE LOGITS
    hradun
    1.20
    Provide
    1.03
    BINDING
    0.98
     ना
    0.97
    Moreover
    0.94
    },[])
    0.93
    It
    0.92
    熾
    0.92
     siempre
    0.92
    CONSTANT
    0.92
    Activations Density 0.000%

    No Known Activations