© 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. 12940
    Prev
    Next
    INDEX
    Explanations

    separator characters after keywords

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

    The neuron is highly sensitive to tokens that name or reference styling classes or style‐related identifiers (e.g. the HTML “class” attribute and CSS/module names).

    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
     in
    -1.56
     proč
    -1.24
    Either
    -1.19
    carian
    -1.19
     either
    -1.14
     what
    -1.10
    勖
    -1.09
    tedly
    -1.08
    ifiable
    -1.08
     generó
    -1.05
    POSITIVE LOGITS
    a
    1.43
    un
    1.30
    ha
    1.29
    '
    1.29
    emph
    1.26
    i
    1.25
     now
    1.23
    j
    1.21
    op
    1.20
    al
    1.16
    Activations Density 0.000%

    No Known Activations