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. 4597
    Prev
    Next
    INDEX
    Explanations

    instances of the word "both" indicating duality or comparison

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

    "both"

    np_acts-logits-general · gemini-2.0-flash

    both X and Y

    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
     vaikka
    -0.73
     qualunque
    -0.71
     magari
    -0.70
    parsedMessage
    -0.66
     zelfs
    -0.64
     архивлан
    -0.63
     addirittura
    -0.62
     orice
    -0.61
     chociaż
    -0.61
     chiunque
    -0.60
    POSITIVE LOGITS
     sides
    0.95
     sexes
    0.82
    sides
    0.77
     Sides
    0.71
     halves
    0.67
    :✨
    0.63
    Sides
    0.61
     ENDS
    0.60
     men
    0.58
    ும்
    0.58
    Activations Density 0.164%

    No Known Activations