© 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. 95849
    Prev
    Next
    INDEX
    Explanations

    closing code punctuation

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

    The neuron fires on closing parenthesis characters—i.e. it detects “)” tokens 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
     sés
    -1.42
    ࿆
    -1.41
     Alamofire
    -1.41
    >>(
    -1.38
    adakan
    -1.33
     veta
    -1.30
    ครื่อง
    -1.28
     avenir
    -1.27
    ၼ်း
    -1.24
    resave
    -1.24
    POSITIVE LOGITS
     pierde
    1.38
     nedeniyle
    1.31
     erfordert
    1.29
    ir
    1.27
     drehen
    1.26
    u
    1.23
     verursacht
    1.22
     brachten
    1.22
     kysy
    1.21
    皷
    1.20
    Activations Density 0.007%

    No Known Activations