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. 2248
    Prev
    Next
    INDEX
    Explanations

    the presence of formatting or structure in the text, such as the beginning of a block or section

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

    Often follows punctuation marks, or square brackets

    np_acts-logits-general · gemini-2.0-flash

    intent / purpose / goal

    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
    principalColumn
    -0.59
    ['
    -0.56
    Referanser
    -0.56
    }}_
    -0.55
    ($_
    -0.54
    ⅘
    -0.52
    ên
    -0.52
    ($"
    -0.51
    WriteLiteral
    -0.51
    $_
    -0.51
    POSITIVE LOGITS
     متعلقه
    0.79
    ']))
    
    0.70
     becauſe
    0.65
     itſelf
    0.64
     ſeveral
    0.63
     !)
    0.61
     ་་
    0.61
     purpoſe
    0.60
    RenderAtEndOf
    0.60
     doubtnut
    0.59
    Activations Density 0.015%

    No Known Activations