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-IT
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 8604
    Prev
    Next
    INDEX
    Explanations

    recurring phrases conveying consistency and reliability

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    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
    MLLoader
    -0.75
     للمعارف
    -0.71
     propOrder
    -0.69
    makeConstraints
    -0.66
    ̈́
    -0.66
     メンテナ
    -0.65
     Weiſe
    -0.65
     deſſen
    -0.64
     好文分享
    -0.64
    awtextra
    -0.63
    POSITIVE LOGITS
     always
    0.99
    always
    0.96
     Always
    0.92
    Always
    0.92
     ALWAYS
    0.89
     siempre
    0.88
     sempre
    0.83
    awsze
    0.81
     zawsze
    0.81
    siempre
    0.79
    Activations Density 0.221%

    No Known Activations