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-2-2B
    4. Residual Stream - 65k
    5. 20-GEMMASCOPE-RES-65K
    6. 504
    Prev
    Next
    INDEX
    Explanations

    references to licenses and limitations in usage

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

    specific expressions or phrases indicating acknowledgment or agreement

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

    names, Okra, slave trade, limitations

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 20-gemmascope-res-65k
    Configuration
    google/gemma-scope-2b-pt-res/layer_20/width_65k/average_l0_114
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    65,536
    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
    =$?
    -0.79
    argout
    -0.67
     '\\;'
    -0.66
    ImageContext
    -0.66
    webElementXpaths
    -0.61
    hoedd
    -0.60
     useContext
    -0.58
    nezeu
    -0.58
    جستارهای
    -0.58
    >';
    
    -0.57
    POSITIVE LOGITS
     rec
    0.53
     initComponents
    0.51
     Sak
    0.49
     presentes
    0.49
     leading
    0.48
     noten
    0.47
     select
    0.47
     war
    0.46
    Bronnen
    0.46
    μισ
    0.46
    Activations Density 0.013%

    No Known Activations