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-IT
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 14507
    Prev
    Next
    INDEX
    Explanations

    specific relationships and roles within familial and scientific contexts

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

    followed by "of" or "to"

    np_acts-logits-general · gemini-2.0-flash

    anticoagulation, family, medical terms

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91
    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
     Horizon
    -0.35
     Thrust
    -0.34
    HOR
    -0.34
     HOR
    -0.33
    ksom
    -0.32
    Hor
    -0.32
    导
    -0.32
     conditionally
    -0.32
     consegue
    -0.32
    Gay
    -0.31
    POSITIVE LOGITS
     ujednoznacz
    0.71
    MessageTagHelper
    0.71
     disambiguazione
    0.66
     estekak
    0.65
    httphttps
    0.61
     GoogleFonts
    0.60
    GeneratedValue
    0.59
    RenderAtEndOf
    0.59
     AssemblyProduct
    0.59
    UnsafeEnabled
    0.57
    Activations Density 0.890%

    No Known Activations