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. 30638
    Prev
    Next
    INDEX
    Explanations

    content related to healthy eating and meal preparation within a family context

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

    healthy dinner recipes

    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
     Paling
    -0.37
     substance
    -0.36
    pezifische
    -0.36
     sztu
    -0.36
    pezif
    -0.36
     Substance
    -0.35
     Clow
    -0.35
     verhält
    -0.35
     szczegół
    -0.35
     Velasco
    -0.34
    POSITIVE LOGITS
     cooking
    0.71
     houſe
    0.66
     purpoſe
    0.65
     cuisines
    0.64
     kasarigan
    0.61
    cooking
    0.57
     pleaſure
    0.57
     meals
    0.56
     occaf
    0.55
     myſelf
    0.55
    Activations Density 0.030%

    No Known Activations