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-2-2B
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 3053
    Prev
    Next
    INDEX
    Explanations

    phrases expressing levels of intensity or degrees of comparison

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

    "bit of" followed by a noun

    np_acts-logits-general · gemini-2.0-flash

    a bit of

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    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
    makeConstraints
    -0.73
     незавершена
    -0.69
    excellents
    -0.62
     Pingback
    -0.62
    ⋙
    -0.61
    Personensuche
    -0.59
    indd
    -0.59
     consumi
    -0.58
     efficacement
    -0.58
     NgModule
    -0.57
    POSITIVE LOGITS
     bit
    0.96
     more
    0.92
    ValueStyle
    0.83
     less
    0.76
     too
    0.74
     longer
    0.72
    bit
    0.69
     BIT
    0.67
    Bit
    0.67
     different
    0.66
    Activations Density 0.175%

    No Known Activations