© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 43849
    Prev
    Next
    INDEX
    Explanations

    verbatim text, verbose printing

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

    The neuron is firing on configuration‐style keywords or flags in code—things like “verbose,” “-v,” “StrictMode,” “verbatim,” etc.—i.e. tokens naming option or mode settings.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
    廕
    -2.48
    МЫ
    -2.36
    aabb
    -2.33
    푢
    -2.28
    ICOLON
    -2.28
     diverso
    -2.27
    攰
    -2.23
     amazingly
    -2.22
    </h5>
    -2.22
     ‘’
    -2.20
    POSITIVE LOGITS
    8
    4.03
    6
    3.95
    7
    3.94
    9
    3.33
    3
    3.30
    5
    3.20
    2
    3.16
    0
    3.06
    .
    3.03
    4
    2.91
    Activations Density 0.000%

    No Known Activations