© 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. 33893
    Prev
    Next
    INDEX
    Explanations

    or and alternatives

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

    The neuron detects delimiter symbols—especially forward‐slashes (and similar punctuation like pipes)—that separate elements in code, URIs, or markup.

    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
     that
    -1.23
     then
    -1.06
     preocupação
    -1.02
     previsão
    -0.96
     opinião
    -0.94
     Conseils
    -0.92
    ammlung
    -0.92
     Inicio
    -0.91
    Ensuite
    -0.91
     resistência
    -0.90
    POSITIVE LOGITS
     or
    1.46
     otros
    1.07
     alternatively
    1.03
    ｖ
    1.00
    or
    1.00
    이나
    0.99
     alternativa
    0.99
     others
    0.99
    other
    0.98
     oder
    0.96
    Activations Density 0.002%

    No Known Activations