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 - 131k
    5. 20-GEMMASCOPE-RES-131K
    6. 14254
    Prev
    Next
    INDEX
    Explanations

    mathematical expressions and equations related to functions and variables

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

    mathematical (`u`, `x`, `cos`) and code (`dir`, `vertical`) tokens

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81
    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
     kasarigan
    -0.54
    webElement
    -0.49
    parsedMessage
    -0.46
    SOUNDBITE
    -0.44
    ViewFeatures
    -0.43
     '{@
    -0.42
    出版年
    -0.41
    httphttps
    -0.41
    GeneratedValue
    -0.41
    GenerationType
    -0.40
    POSITIVE LOGITS
    lapsingToolbar
    0.39
    yaret
    0.38
     lus
    0.34
     capitolo
    0.34
    __.__
    0.34
    his
    0.33
     Traps
    0.33
    паси
    0.33
    bro
    0.33
    ыл
    0.32
    Activations Density 0.059%

    No Known Activations