© 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
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 13595
    Prev
    Next
    INDEX
    Explanations

    section markers and separators

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

    The neuron strongly activates on section‐heading markup tokens (e.g. “SH” or “section”) that mark the start of a new document section.

    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
    に入れる
    -1.48
     também
    -1.41
    帰る
    -1.38
    の方も
    -1.33
    でない
    -1.32
    も行
    -1.30
    状態で
    -1.25
     และ
    -1.23
    に入る
    -1.19
     וכן
    -1.17
    POSITIVE LOGITS
    (
    1.41
    %">
    1.34
    iligten
    1.24
     denominada
    1.22
    %;">
    1.22
     strå
    1.20
    !!!!
    1.18
    <>
    
    1.17
    .’’
    1.16
    »
    1.16
    Activations Density 0.021%

    No Known Activations