© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 25174
    Prev
    Next
    INDEX
    Explanations

    technical documentation formatting

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

    The neuron is primarily sensitive to section‐header tokens—e.g. markdown “#”/“##” markers and front-matter labels or other heading words (like “The,” “References,” “Namespace,” “title”) that introduce new document sections.

    oai_token-act-pair · o4-mini
    Triggered 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
    Negative Logits
     its
    -0.93
     most
    -0.88
    eql
    -0.79
    Ղ
    -0.78
     Gegenstand
    -0.76
    pezi
    -0.76
     any
    -0.75
    offsetHeight
    -0.75
     exist
    -0.75
    compareTo
    -0.75
    POSITIVE LOGITS
    😒
    0.85
     gravelly
    0.81
     сті
    0.81
    的实力
    0.81
    witness
    0.80
     liquids
    0.79
    
    0.79
    hanger
    0.77
    関連リンク
    0.77
     Fertility
    0.76
    Activations Density 0.021%

    No Known Activations