© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 129803
    Prev
    Next
    INDEX
    Explanations

    The neuron detects curly-brace tokens (“{” and “}”) that mark the start and end of code blocks.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    list, change, create, remove

    np_acts-logits-general · gemini-2.5-flash-lite
    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
    lewati
    -0.96
    篭
    -0.92
     hectáreas
    -0.90
    苛
    -0.90
    opencv
    -0.89
     égard
    -0.88
     svet
    -0.88
    WithDefault
    -0.88
     surrounded
    -0.88
     Overall
    -0.88
    POSITIVE LOGITS
    接下來
    0.93
    そこから
    0.92
    一个个
    0.92
     poszczegól
    0.90
     faciliter
    0.88
    臉書
    0.87
    認真
    0.86
    灑
    0.86
     einzelne
    0.85
     pouze
    0.84
    Activations Density 0.008%

    No Known Activations