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. 20-GEMMASCOPE-RES-65K
    6. 976
    Prev
    Next
    INDEX
    Explanations

    numerical representations or mathematical expressions

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

    numerical values or quantities in a mathematical context

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

    numbers and mathematical expressions

    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-65k
    Configuration
    google/gemma-scope-2b-pt-res/layer_20/width_65k/average_l0_114
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    65,536
    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
    GEBURTSDATUM
    -0.96
    ✭✭
    -0.92
    Hozzáférés
    -0.91
    OGND
    -0.89
    ModelAdmin
    -0.86
    reportWebVitals
    -0.86
    SaveChangesAsync
    -0.85
    ChildScrollView
    -0.85
    Lähteet
    -0.84
    CodedInputStream
    -0.84
    POSITIVE LOGITS
    ing
    0.52
    ning
    0.51
    bbero
    0.47
    5
    0.46
    4
    0.46
    ValueStyle
    0.44
    7
    0.42
    usk
    0.41
    Rig
    0.41
    deg
    0.41
    Activations Density 0.187%

    No Known Activations