© 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. 22-GEMMASCOPE-RES-131K
    7. 12788
    Prev
    Next
    INDEX
    Explanations

    scattered all, these materials, master alike

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

    This neuron selectively fires on code‐style tokens—especially import/package/includes, macro/define lines, and other programming-identifier keywords.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
    fani
    -0.78
     recollect
    -0.77
     ModelExpression
    -0.75
     Offisielt
    -0.73
    MEK
    -0.73
    ⬎
    -0.71
    分手
    -0.71
     underage
    -0.71
    IPPI
    -0.71
     reimburse
    -0.70
    POSITIVE LOGITS
     osu
    0.88
     tè
    0.84
    érations
    0.83
     Gea
    0.82
    濘
    0.82
     phytoplankton
    0.81
    omorphisms
    0.80
     Himal
    0.80
     dji
    0.80
     Trout
    0.79
    Activations Density 0.276%

    No Known Activations