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 - 32k
    5. 12-GEMMASCOPE-RES-32K
    6. 1727
    Prev
    Next
    INDEX
    Explanations

    terms related to sealing and leakage prevention

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2b-pt-res/layer_12/width_32k/average_l0_76
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    32,768
    Data Type
    float32
    Hook Name
    blocks.12.hook_resid_post
    Hook Layer
    12
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    Билгалдахарш
    -0.59
    resiasi
    -0.55
     fusible
    -0.53
     Hap
    -0.51
    LOADED
    -0.51
     aloud
    -0.50
    orylation
    -0.49
    RUnlock
    -0.47
     astore
    -0.46
    UrlResolution
    -0.46
    POSITIVE LOGITS
     sealing
    0.71
    webElementGuid
    0.67
     rapat
    0.63
    IntoConstraints
    0.61
     airtight
    0.61
    CodedInputStream
    0.60
    MENAFN
    0.57
     Sealing
    0.57
     виправивши
    0.57
     seal
    0.56
    Activations Density 0.227%

    No Known Activations