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. 17-GEMMASCOPE-RES-65K
    6. 28456
    Prev
    Next
    INDEX
    Explanations

    specific subjects and significant issues within various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 17-gemmascope-res-65k
    Configuration
    google/gemma-scope-2b-pt-res/layer_17/width_65k/average_l0_125
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    65,536
    Data Type
    float32
    Hook Name
    blocks.17.hook_resid_post
    Hook Layer
    17
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    تقاوى
    -0.63
    fromCharCode
    -0.59
    WireFormatLite
    -0.59
    GraphicsUnit
    -0.53
     Мексичка
    -0.53
    ModelAttribute
    -0.53
    ApplicationTests
    -0.52
    posedge
    -0.52
     nakalista
    -0.52
     пунктов
    -0.51
    POSITIVE LOGITS
     into
    1.13
     onto
    0.98
     Into
    0.88
    Into
    0.85
    into
    0.77
     away
    0.76
     INTO
    0.75
     down
    0.73
     out
    0.72
     off
    0.59
    Activations Density 0.313%

    No Known Activations