© 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. 63084
    Prev
    Next
    INDEX
    Explanations

    code separators and keywords

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

    The neuron responds to long all‐caps code identifiers or macro names, especially those containing underscores.

    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
    anas
    -0.92
    man
    -0.91
    â
    -0.91
    *}\
    -0.90
     $^{
    -0.90
     regimen
    -0.87
    —"
    -0.86
    dée
    -0.85
    ‐
    -0.85
    and
    -0.85
    POSITIVE LOGITS
    Répondre
    0.97
    правление
    0.97
    ادامه
    0.93
    Réalisation
    0.93
    읖
    0.88
     illustrationer
    0.88
    の地
    0.87
     angene
    0.87
    natured
    0.86
    Citiți
    0.86
    Activations Density 0.188%

    No Known Activations