© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 1040
    Prev
    Next
    INDEX
    Explanations

    EF, CLR, names, and embedding

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

    The neuron fires on programming‐language tokens (identifiers, keywords, and punctuation in a code snippet) rather than on ordinary prose.

    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
    ينة
    -0.81
    berfläche
    -0.80
    illot
    -0.78
    hibli
    -0.72
     Oatmeal
    -0.70
    VALID
    -0.70
     ఉన్న
    -0.69
    RNA
    -0.69
    атра
    -0.68
    scrape
    -0.68
    POSITIVE LOGITS
     EF
    0.94
     CLR
    0.78
    ：(
    0.75
    clr
    0.73
    ：</
    0.73
    tamente
    0.72
     Gier
    0.72
     secretaria
    0.71
     Volo
    0.71
     nunca
    0.71
    Activations Density 0.020%

    No Known Activations