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. 56613
    Prev
    Next
    INDEX
    Explanations

    terms that express possession or connection to an object or subject

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

    water elemental, strict, people, Somerset, bull, rich

    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
    setVerticalGroup
    -0.63
    enterOuterAlt
    -0.46
     OMITBAD
    -0.40
     Numerade
    -0.37
    __':
    -0.36
    __':
    
    -0.36
    MouseClicked
    -0.36
    جغرافيا
    -0.34
    OuterAlt
    -0.34
    '])){
    -0.34
    POSITIVE LOGITS
    ValueStyle
    0.63
    PerformLayout
    0.60
    RenderAtEndOf
    0.54
     informée
    0.53
    pholes
    0.52
     esternos
    0.52
    ratto
    0.52
    HasAnnotation
    0.51
    enderror
    0.51
     Judea
    0.51
    Activations Density 0.056%

    No Known Activations