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. 4058
    Prev
    Next
    INDEX
    Explanations

    references to specific measurements or experimental data in scientific contexts

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

    Numbers followed by parentheses and a file extension

    np_acts-logits-general · gemini-2.0-flash

    scaling and size-related

    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
    AndEndTag
    -0.97
    stateProvider
    -0.85
     unknownFields
    -0.84
     CreateTagHelper
    -0.83
    :✨
    -0.83
    __":
    
    -0.82
     ModelExpression
    -0.80
     ویکی‌پدیای
    -0.78
     '\\;'
    -0.77
    ']))
    
    -0.76
    POSITIVE LOGITS
     enlarged
    0.68
     enlargement
    0.63
     zoomed
    0.63
     zooming
    0.61
     resized
    0.60
     enlarging
    0.60
    _
    0.59
     resizing
    0.59
     Enlarge
    0.57
     Plage
    0.55
    Activations Density 0.534%

    No Known Activations