© 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
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 31246
    Prev
    Next
    INDEX
    Explanations

    Configuration, resources, and code structures

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

    The neuron fires on high-utility terms in code comments or auto-generated documentation blocks (e.g. “The”, “main”, “entry”, “point”, “for”, “application”, “used”, etc.), i.e. it detects descriptive header or metadata text in source‐code 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_34/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
     presidencial
    -0.91
     sindic
    -0.85
    .~(\
    -0.82
     convento
    -0.80
     focusing
    -0.79
    场比赛
    -0.78
     hispan
    -0.78
    يدي
    -0.77
     Distrito
    -0.75
     considerable
    -0.75
    POSITIVE LOGITS
     resourceCulture
    1.09
     strings
    0.97
     string
    0.89
    ulturen
    0.87
     cached
    0.84
    字符串
    0.83
    oar
    0.80
    insatz
    0.80
     resource
    0.79
     Resources
    0.79
    Activations Density 0.001%

    No Known Activations