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. 20-GEMMASCOPE-RES-16K
    6. 253
    Prev
    Next
    INDEX
    Explanations

    phrases related to historical context and events involving power dynamics and human experiences

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

    inflammatory or conspiratorial rhetoric about societal power structures and systemic oppression.

    oai_token-act-pair · claude-3-5-haiku-20241022Triggered by @neilrathi
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_20/width_16k/average_l0_68
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    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
    قایناق‌لار
    -0.63
     disambiguazione
    -0.62
     ویکی‌پدیا
    -0.61
    цездатний
    -0.61
     ProtoMessage
    -0.60
    rungsseite
    -0.60
    verwijspagina
    -0.60
    Tembelea
    -0.59
    makeConstraints
    -0.58
    CppCodeGen
    -0.57
    POSITIVE LOGITS
     absolutely
    0.49
     RIPRODUZIONE
    0.44
     every
    0.44
     badass
    0.43
     forever
    0.43
     навсегда
    0.42
     freakin
    0.42
     instantly
    0.41
     абсолютно
    0.41
    !
    0.40
    Activations Density 0.651%

    No Known Activations