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. 113739
    Prev
    Next
    INDEX
    Explanations

    mentions of family members, especially mothers and parents

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    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
    Datuak
    -0.40
     Ze
    -0.38
    gdx
    -0.36
     veremos
    -0.36
     hand
    -0.35
    Misc
    -0.35
    съ
    -0.35
    BorderLayout
    -0.35
    fla
    -0.35
     mis
    -0.35
    POSITIVE LOGITS
     father
    0.66
     grandfather
    0.58
     Father
    0.58
    👵
    0.58
     grandparents
    0.58
     Grandfather
    0.57
     FATHER
    0.57
     abuelo
    0.56
     grandmother
    0.55
     abuela
    0.54
    Activations Density 0.044%

    No Known Activations