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. 39504
    Prev
    Next
    INDEX
    Explanations

    terms related to supplementary materials and dietary regulations

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

    supplementary and '-mentary' words

    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
     feroit
    -0.73
     ainfi
    -0.71
     pouvoit
    -0.69
     étoit
    -0.68
     auroit
    -0.63
     Houſe
    -0.61
    følgelig
    -0.60
     gobern
    -0.60
     avoient
    -0.59
    näm
    -0.59
    POSITIVE LOGITS
    neurial
    0.81
    ratory
    0.81
    ductory
    0.79
    etary
    0.77
    atory
    0.68
    ulatory
    0.66
    mentary
    0.66
    lammatory
    0.66
    utory
    0.65
    piratory
    0.64
    Activations Density 0.046%

    No Known Activations