© 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
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 50328
    Prev
    Next
    INDEX
    Explanations

    .m, .h, .swift

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

    The neuron fires strongly on the “.” that separates a source‐code filename from its extension (e.g. “Context .m”, “MediaObserver .swift”, “Utils .h”).

    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
     there
    -1.17
     in
    -0.99
    ;
    -0.91
     on
    -0.87
    tsx
    -0.82
     бли
    -0.80
     In
    -0.79
     without
    -0.79
     when
    -0.77
    彗
    -0.76
    POSITIVE LOGITS
    cpp
    1.13
    Öl
    0.95
     pensamientos
    0.94
    Folgende
    0.93
     coûts
    0.93
     énorme
    0.92
     کنم
    0.91
    臃
    0.91
     которые
    0.91
     supprimer
    0.90
    Activations Density 0.010%

    No Known Activations