Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-2B
    4. Residual Stream - 16k
    5. 18-GEMMASCOPE-RES-16K
    6. 2174
    Prev
    Next
    INDEX
    Explanations

    references to fruit and related products

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

    "fruit" or its variations

    np_acts-logits-general · gemini-2.0-flash

    fruit or fruits

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.18.hook_resid_post
    Hook Layer
    18
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    addOn
    -0.47
    nd
    -0.45
    glfw
    -0.44
    ensatz
    -0.43
     sest
    -0.43
     ход
    -0.42
    ("${
    -0.42
    ρε
    -0.42
    machte
    -0.42
     fire
    -0.41
    POSITIVE LOGITS
     fruit
    1.28
     fruits
    1.27
     Fruit
    1.21
     Fruits
    1.21
     orchards
    1.19
     FRUIT
    1.18
    Fruits
    1.14
    Fruit
    1.12
    fruit
    1.09
     orchard
    1.07
    Activations Density 0.173%

    No Known Activations