Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B-IT
    4. Residual Stream - 131k
    5. 20-GEMMASCOPE-RES-131K
    6. 86499
    Prev
    Next
    INDEX
    Explanations

    references to social etiquette and norms in behavior

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    social rules and etiquette

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    131,072
    Data Type
    float32
    Hook Name
    blocks.20.hook_resid_post
    Hook Layer
    20
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    rboles
    -0.61
    GtkWidget
    -0.54
    ifolia
    -0.52
    __':
    
    -0.51
    ctrica
    -0.50
    yelitis
    -0.49
    //
    -0.48
    colnshire
    -0.48
    MultipartFile
    -0.47
    ChildScrollView
    -0.47
    POSITIVE LOGITS
     etiquette
    0.76
     Etiquette
    0.73
     rules
    0.58
     reglas
    0.56
    rules
    0.54
    Rules
    0.54
     Rules
    0.52
     regler
    0.52
     conventions
    0.50
     regels
    0.49
    Activations Density 0.058%

    No Known Activations