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-2-9B
    4. Residual Stream - 16k
    5. 21-GEMMASCOPE-RES-16K
    6. 2135
    Prev
    Next
    INDEX
    Explanations

    phrases indicating conditional or time-related scenarios

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 21-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_21/width_16k/average_l0_129
    Prompts (Dashboard)
    24,576 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
    Itoa
    -0.45
    logError
    -0.44
     result
    -0.44
     getLayout
    -0.42
     Steele
    -0.42
    SizeF
    -0.41
    もしれません
    -0.39
     error
    -0.39
    er
    -0.38
    nehå
    -0.37
    POSITIVE LOGITS
    when
    1.22
     WHEN
    1.21
    Cuándo
    1.18
     when
    1.14
    WHEN
    1.11
    cuando
    1.11
     Cuando
    1.09
     When
    1.09
     Quando
    1.08
    When
    1.07
    Activations Density 0.195%

    No Known Activations