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

    Neuronpedia

    APIAssistant AxisNEWCircuit 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. 45863
    Prev
    Next
    INDEX
    Explanations

    justifications and legal reasoning behind actions or decisions

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

    justified since, was a, that the

    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
    ността
    -0.48
    ARM
    -0.40
    mely
    -0.40
    DoubleQuotes
    -0.39
    arma
    -0.38
    titleMargin
    -0.38
    ensor
    -0.38
    hist
    -0.37
    tome
    -0.37
     clou
    -0.36
    POSITIVE LOGITS
     समीक्षाओं
    0.47
     للاسماء
    0.47
    featureID
    0.47
    Rohy
    0.45
    AnchorTagHelper
    0.40
    LabelTagHelper
    0.39
    IndentedString
    0.39
    )_/¯
    0.39
     bluzka
    0.39
     rationale
    0.38
    Activations Density 0.081%

    No Known Activations