© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B
    4. Residual Stream - 16k
    5. 38-GEMMASCOPE-RES-16K
    6. 7943
    Prev
    Next
    INDEX
    Explanations

    technical terms and conditions related to mechanics and electronics

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 38-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_38/width_16k/average_l0_128
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.38.hook_resid_post
    Hook Layer
    38
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     CanadaChoose
    -0.53
    BeginContext
    -0.53
    RenderAtEndOf
    -0.52
     transfieras
    -0.48
     ujednoznacz
    -0.46
    Parcelize
    -0.45
    出版年
    -0.44
    MLLoader
    -0.44
    ніципалі
    -0.44
    +#+#
    -0.43
    POSITIVE LOGITS
     closed
    3.53
     closing
    3.25
     closes
    3.09
     Closed
    3.02
     closure
    3.02
    closed
    2.91
     Closing
    2.88
    Closed
    2.83
     clos
    2.77
    Closing
    2.70
    Activations Density 1.158%

    No Known Activations