© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26
    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 98705
    Prev
    Next
    INDEX
    Explanations

    ifndef and define

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

    The neuron fires on identifier fragments that end with an underscore (e.g. library or macro name prefixes like “ngx_”, “ASIO_”, “Py_”, “MPI_”, etc.).

    oai_token-act-pair · o4-mini
    Triggered 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
    Negative Logits
    1
    -1.99
    UGHTER
    -1.58
    さらに
    -1.55
     Many
    -1.53
     Following
    -1.44
    Many
    -1.43
     for
    -1.41
     While
    -1.38
     including
    -1.38
    .,
    -1.37
    POSITIVE LOGITS
    </i>
    1.87
    ceased
    1.77
     riscos
    1.75
    甍
    1.66
    🫥
    1.65
     probleem
    1.60
     concreta
    1.60
     nativa
    1.59
    大丈夫です
    1.58
    ⡑
    1.58
    Activations Density 0.019%

    No Known Activations