Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & 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. 33-GEMMASCOPE-RES-16K
    6. 426
    Prev
    Next
    INDEX
    Explanations

    references to the concept of "magic."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 33-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_33/width_16k/average_l0_114
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.33.hook_resid_post
    Hook Layer
    33
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    帖最后由
    -0.38
    TRAVIS
    -0.37
     stolz
    -0.37
     пунктов
    -0.36
     kiin
    -0.36
    ureusement
    -0.35
     Anlagen
    -0.35
     primas
    -0.35
     확인함
    -0.35
     arşivlendi
    -0.35
    POSITIVE LOGITS
     Magic
    0.71
     MAGIC
    0.67
    Magic
    0.65
     magic
    0.64
    🪄
    0.62
     wand
    0.58
    magic
    0.58
    MAGIC
    0.57
     trick
    0.55
     Magical
    0.55
    Activations Density 0.173%

    No Known Activations