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 - 16k
    5. 21-GEMMASCOPE-RES-16K
    6. 605
    Prev
    Next
    INDEX
    Explanations

    instances of HTML tags and attributes

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 21-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_21/width_16k/average_l0_70
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.21.hook_resid_post
    Hook Layer
    21
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ($__
    -0.52
     مشين
    -0.50
    subplots
    -0.48
    TemporalType
    -0.48
    SpringRunner
    -0.47
    ({},
    -0.46
     كومونز
    -0.46
     front
    -0.45
    SharedCtor
    -0.45
    offsetof
    -0.43
    POSITIVE LOGITS
    <strong>
    0.68
     CreateTagHelper
    0.68
     ProtoMessage
    0.67
    <b>
    0.65
    tagHelperRunner
    0.62
    MLLoader
    0.62
    batore
    0.58
    modelBuilder
    0.56
    ykjav
    0.54
    ScopeManager
    0.54
    Activations Density 0.095%

    No Known Activations