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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsBlogSlackPrivacy & 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. 12811
    Prev
    Next
    INDEX
    Explanations

    instances of the word "dance" and its variations, indicating a focus on dance-related activities and contexts

    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
    roek
    -0.47
     bemerk
    -0.39
    "}},
    -0.39
    ()].
    -0.39
    worfen
    -0.39
     fik
    -0.38
     Gunner
    -0.38
    mlung
    -0.37
     lende
    -0.36
    )">
    -0.36
    POSITIVE LOGITS
     dance
    1.57
     DANCE
    1.39
     Dance
    1.38
    Dance
    1.32
     danced
    1.29
     dancing
    1.28
    dance
    1.28
     dancers
    1.20
     dances
    1.18
     dancer
    1.13
    Activations Density 0.124%

    No Known Activations