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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B
    4. Residual Stream - 131k
    5. 20-GEMMASCOPE-RES-131K
    6. 53907
    Prev
    Next
    INDEX
    Explanations

    instances of the word "though" and its variations, indicating contrast or concession in discourse

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114
    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
     ویکی‌پدی
    -0.65
    PMailer
    -0.56
    %")
    -0.55
    )}</
    -0.55
     invokingState
    -0.49
     BPI
    -0.49
    %");
    -0.49
    ChildScrollView
    -0.49
    ksikon
    -0.49
     getCity
    -0.48
    POSITIVE LOGITS
     THOUGH
    0.66
     though
    0.62
    ,
    0.58
     Anyway
    0.56
    though
    0.56
     But
    0.56
    Anyway
    0.56
    Anyhow
    0.54
    But
    0.53
     Though
    0.53
    Activations Density 0.007%

    No Known Activations