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. 15-GEMMASCOPE-RES-16K
    6. 1825
    Prev
    Next
    INDEX
    Explanations

    the presence of significant structural markers in a text, such as headers or sections

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 15-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_15/width_16k/average_l0_78
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.15.hook_resid_post
    Hook Layer
    15
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     trekken
    -0.68
     isolato
    -0.59
     représenté
    -0.58
     hunne
    -0.58
    libft
    -0.57
     valable
    -0.56
     réun
    -0.55
    Forrás
    -0.54
     mierda
    -0.53
     cuchillo
    -0.53
    POSITIVE LOGITS
    Title
    0.81
     مشين
    0.75
    \{\\
    0.74
    awtextra
    0.74
     estimés
    0.69
     ***!
    0.67
     Title
    0.67
     Савезне
    0.62
     CreateTagHelper
    0.60
    djangoproject
    0.59
    Activations Density 0.007%

    No Known Activations