© 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. 70609
    Prev
    Next
    INDEX
    Explanations

    role and place in society

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

    The neuron detects occurrences of section‐style headings that follow the pattern “Role (or Roles) of … in …,” i.e. titles stating the role of something in some context.

    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
     werdet
    -1.06
     conservé
    -1.05
     방법
    -1.03
     passiert
    -0.97
    洹
    -0.96
    atorial
    -0.95
    Produit
    -0.95
     sandalias
    -0.94
    本部
    -0.93
    
    -0.93
    POSITIVE LOGITS
     role
    1.15
     in
    1.12
     роль
    1.04
     as
    1.01
    ในการ
    0.99
     society
    0.94
     relation
    0.93
     согласно
    0.92
    role
    0.91
     после
    0.91
    Activations Density 0.086%

    No Known Activations