© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 120095
    Prev
    Next
    INDEX
    Explanations

    copyright and license headers

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

    long sequences of dashes used as visual separators in code comments and headers.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @bfgcn7frf5

    The neuron detects long horizontal separator lines made up of repeated dash characters in code or comments.

    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
    小行星
    -1.54
     بسبب
    -1.49
    私が
    -1.36
    Алексей
    -1.35
    sprüng
    -1.34
    私も
    -1.33
    !"
    -1.32
    ⇲
    -1.30
    Kesimpulan
    -1.30
    Какие
    -1.30
    POSITIVE LOGITS
     that
    2.00
     which
    1.85
     and
    1.84
     of
    1.83
     until
    1.80
     including
    1.62
    </i>
    1.50
     at
    1.49
     to
    1.48
     they
    1.48
    Activations Density 0.037%

    No Known Activations