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 - 65k
    5. 16-GEMMASCOPE-RES-65K
    6. 52949
    Prev
    Next
    INDEX
    Explanations

    the presence of structured data or coding syntax

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

    to settle

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 16-gemmascope-res-65k
    Configuration
    google/gemma-scope-2b-pt-res/layer_16/width_65k/average_l0_128
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    65,536
    Data Type
    float32
    Hook Name
    blocks.16.hook_resid_post
    Hook Layer
    16
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Tsar
    -0.43
     comigo
    -0.42
     duchess
    -0.40
     Duchess
    -0.40
    əsi
    -0.39
     hierogly
    -0.39
     Suomessa
    -0.38
     isotopic
    -0.37
     econô
    -0.36
     gezin
    -0.36
    POSITIVE LOGITS
    مصادر
    0.95
     مشين
    0.92
    UnusedPrivate
    0.79
     transfieras
    0.74
    httphttps
    0.69
    parsedMessage
    0.69
    oa̍t
    0.69
    Geplaatst
    0.67
    ագրություններ
    0.67
    ViewImports
    0.66
    Activations Density 0.512%

    No Known Activations