© 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. 54121
    Prev
    Next
    INDEX
    Explanations

    questions or code snippets

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

    The neuron spikes on bracket‐like delimiters (e.g. parentheses, angle brackets, braces) that mark grouped or parenthesized text.

    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
    меты
    -1.15
     нѣ
    -1.13
     billeder
    -1.11
     forskellige
    -1.09
     Пен
    -1.09
     jaunes
    -1.09
    ày
    -1.08
    푼
    -1.08
     obrázok
    -1.08
    Ϟ
    -1.07
    POSITIVE LOGITS
     when
    1.17
     while
    1.16
     going
    1.05
     Then
    1.04
     where
    1.04
     can
    1.01
     making
    1.01
    떄
    1.00
     While
    0.99
     aiming
    0.98
    Activations Density 0.019%

    No Known Activations