© 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-IT
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 676
    Prev
    Next
    INDEX
    Explanations

    inquiries or questions related to methodology and processes

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

    "how"

    np_acts-logits-general · gemini-2.0-flash

    how to or how much

    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
     element
    -0.36
    WidgetItem
    -0.36
     present
    -0.36
     subject
    -0.35
    Willi
    -0.35
     Eldridge
    -0.35
     있다
    -0.35
     Deane
    -0.35
    십시오
    -0.34
     purpo
    -0.33
    POSITIVE LOGITS
     how
    1.38
    How
    1.29
     How
    1.25
     HOW
    1.16
    how
    1.14
    HOW
    1.05
    Cómo
    0.96
     cómo
    0.96
    howto
    0.91
     Howe
    0.87
    Activations Density 0.059%

    No Known Activations