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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B-IT
    4. Residual Stream - 131k
    5. 20-GEMMASCOPE-RES-131K
    6. 78901
    Prev
    Next
    INDEX
    Explanations

    specific HTML tags or elements in code, particularly `div` and `script` tags

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

    JavaScript snippets and tests

    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-131k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    131,072
    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
     CreateTagHelper
    -0.67
     zijne
    -0.64
     indígen
    -0.59
    uxxxx
    -0.58
    rzost
    -0.57
    :+:
    -0.55
    tonode
    -0.54
    msgTypes
    -0.54
     increí
    -0.54
     queſta
    -0.54
    POSITIVE LOGITS
    ChildScrollView
    0.48
     MaterialApp
    0.47
    verwijspagina
    0.44
    VStack
    0.40
     utafitiHapana
    0.39
    setVerticalGroup
    0.39
    CommonModule
    0.37
    IComponent
    0.35
    stdio
    0.35
    TE
    0.34
    Activations Density 0.001%

    No Known Activations