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. 8185
    Prev
    Next
    INDEX
    Explanations

    pieces of code and programming constructs related to APIs and JSON handling

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

    Tokens appearing before code snippets

    np_acts-logits-general · gemini-2.0-flash

    into, a, an

    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
     <=",
    -0.60
     Theſe
    -0.51
     beſch
    -0.51
     مشين
    -0.50
     contextLoads
    -0.49
     becauſe
    -0.48
    Hauptartikel
    -0.47
    chränk
    -0.46
     ligiloj
    -0.46
     BrowserModule
    -0.45
    POSITIVE LOGITS
     BoxFit
    0.43
    WithMany
    0.41
    adpleegd
    0.35
    URLException
    0.34
    代わりに
    0.34
    -------------</
    0.34
    MessageTagHelper
    0.33
    RenderAtEndOf
    0.33
     doc
    0.33
    Runner
    0.33
    Activations Density 0.924%

    No Known Activations