© 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
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 130686
    Prev
    Next
    INDEX
    Explanations

    The neuron is triggered by parenthetical or bracketed text markers (e.g. “(”, “)”, “[”, or “]”) indicating asides or inline citations.

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

    closing brackets and parentheses

    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
    というわけで
    -1.85
    Jangan
    -1.46
    <bos>
    -1.42
    それにしても
    -1.40
    ちなみに
    -1.36
    日時
    -1.35
    そういえば
    -1.35
    Waktu
    -1.34
    <0xA3>
    -1.32
    کفش
    -1.30
    POSITIVE LOGITS
     amazing
    1.54
    逮
    1.52
     extremely
    1.41
     soooo
    1.38
     fabulous
    1.38
     exceedingly
    1.38
    ってる
    1.36
     astounding
    1.35
     insanely
    1.35
     amazingly
    1.34
    Activations Density 0.002%

    No Known Activations