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. 627
    Prev
    Next
    INDEX
    Explanations

    variations of the pronoun "you" and punctuation marks indicating emphasis or dialogue

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

    quotation marks or apostrophes indicating dialogue or possession

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

    numerical sequences, equations, times

    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
    ]<=
    -0.67
    GEBURTSDATUM
    -0.65
    ]>=
    -0.65
    "]=
    -0.64
    $",
    -0.63
     nos
    -0.61
     {}",
    -0.60
    =>'
    -0.60
    </th>
    -0.59
    WebServlet
    -0.57
    POSITIVE LOGITS
     Although
    0.88
    Although
    0.88
     While
    0.85
    While
    0.85
    There
    0.83
     There
    0.82
     However
    0.80
     Some
    0.79
     During
    0.79
    However
    0.78
    Activations Density 0.092%

    No Known Activations