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. 6686
    Prev
    Next
    INDEX
    Explanations

    references to "the" and related articles, suggesting an emphasis on specificity or particularity in the context presented

    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
     nemlig
    -0.82
     totiž
    -0.75
     namelijk
    -0.68
     >=",
    -0.66
     nämlich
    -0.64
     either
    -0.64
     entweder
    -0.63
    either
    -0.61
     wikipagina
    -0.61
     firstly
    -0.60
    POSITIVE LOGITS
     consiguiente
    1.48
     consequent
    1.37
     subsequent
    1.27
     daraus
    1.17
     subsequently
    1.16
     accompanying
    1.15
     consequently
    1.13
     associated
    1.13
    それに
    1.13
    何より
    1.13
    Activations Density 0.437%

    No Known Activations