© 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. 28193
    Prev
    Next
    INDEX
    Explanations

    research, study, and design

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

    The neuron fires most strongly on formal section headings or titles of scientific studies and reports.

    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
     for
    -1.02
     ALL
    -0.81
    不管是
    -0.79
     Fehl
    -0.78
    çoivent
    -0.76
     Gegend
    -0.75
    剧本
    -0.75
     соответственно
    -0.75
     more
    -0.74
     bruising
    -0.73
    POSITIVE LOGITS
    hels
    0.99
    angerie
    0.98
    SGD
    0.97
    vito
    0.96
     léka
    0.93
     verhind
    0.92
    uct
    0.89
    zené
    0.89
    -|-
    0.88
    ciens
    0.88
    Activations Density 0.099%

    No Known Activations