© 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. 130970
    Prev
    Next
    INDEX
    Explanations

    The neuron fires on tokens that are technical or “code‐like” identifiers—e.g. programming function names, XML/HTML tags, scientific-name abbreviations, version numbers, etc.—rather than ordinary running text words.

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

    `ast` followed by `ala` or `()` followed by `ComboBox`

    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
     البته
    -0.99
    ᅠ
    -0.88
     ventilador
    -0.86
     $(-
    -0.84
    Зна
    -0.84
     WITH
    -0.84
    tabilis
    -0.83
    较高
    -0.83
     both
    -0.83
     Kasım
    -0.83
    POSITIVE LOGITS
    }$​
    0.92
     ArrayAdapter
    0.90
    を使って
    0.85
    ську
    0.85
    his
    0.84
    9
    0.84
    ESTADO
    0.84
    皷
    0.83
     Wohn
    0.82
    tawesome
    0.82
    Activations Density 0.015%

    No Known Activations