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. 2311
    Prev
    Next
    INDEX
    Explanations

    mentions of writing instruments and their related materials

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

    pencil, iodine, graphite, crayon

    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
     total
    -0.43
    RemoveField
    -0.41
    resave
    -0.40
     state
    -0.40
    Viited
    -0.39
     rate
    -0.38
     Melrose
    -0.38
     Rivas
    -0.38
     tour
    -0.38
     umsch
    -0.37
    POSITIVE LOGITS
     Pencil
    1.33
    Pencil
    1.27
     pencil
    1.22
    pencil
    1.20
     pencils
    1.13
    ENCIL
    0.93
     каранда
    0.93
    cils
    0.73
     lápiz
    0.73
    鉛筆
    0.72
    Activations Density 0.004%

    No Known Activations