Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit 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. 24-GEMMASCOPE-RES-16K
    6. 4466
    Prev
    Next
    INDEX
    Explanations

    references to curcumin, a compound found in turmeric, and its various forms or uses

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

    starting with "Cur" followed by a letter

    np_acts-logits-general · gemini-2.0-flash

    cur words: Curcuma, curatively

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 24-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_24/width_16k/average_l0_73
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.24.hook_resid_post
    Hook Layer
    24
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    UnsafeEnabled
    -0.65
    IBOutlet
    -0.64
     initComponents
    -0.59
     Anlage
    -0.56
    Демографія
    -0.52
    ambique
    -0.51
     layui
    -0.50
    epam
    -0.49
    annica
    -0.49
     الاطلاع
    -0.48
    POSITIVE LOGITS
     cur
    0.64
     Cur
    0.58
     CUR
    0.56
    tailed
    0.56
    cur
    0.55
    tail
    0.55
    rennt
    0.53
     Vitae
    0.52
    vature
    0.52
     Curran
    0.51
    Activations Density 0.108%

    No Known Activations