© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 14412
    Prev
    Next
    INDEX
    Explanations

    C code syntax and includes

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

    The neuron strongly activates on C “#include” header identifiers (e.g. string, memory, errno, regex, ctype) and also on the literal zero/‘\0’, so in effect it’s detecting C standard‐library header names and null/zero literals.

    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
    絝
    -1.86
    </em>
    -1.82
     ktorí
    -1.70
    гий
    -1.69
     CONSIDERATIONS
    -1.62
    ONESIA
    -1.59
     zapal
    -1.59
    珦
    -1.59
    bluza
    -1.58
     legjobb
    -1.57
    POSITIVE LOGITS
    ?
    1.98
    </h1>
    1.96
    ’
    1.91
    e
    1.65
    elif
    1.58
    \
    1.53
    violet
    1.48
     imma
    1.48
    !
    1.46
    ++;
    1.40
    Activations Density 0.003%

    No Known Activations