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 - 131k
    5. 20-GEMMASCOPE-RES-131K
    6. 29593
    Prev
    Next
    INDEX
    Explanations

    occurrences of specific sequence patterns or formats within data strings or code

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

    numerical/symbolic lists

    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
    ロウィン
    -0.58
     pleaſure
    -0.56
     Infór
    -0.56
     tecnológica
    -0.56
     Artículos
    -0.55
     automatiques
    -0.54
     meditación
    -0.54
     navideño
    -0.54
     desmotivaciones
    -0.54
    usermodel
    -0.53
    POSITIVE LOGITS
     estekak
    0.49
    setVerticalGroup
    0.43
    ViewFeatures
    0.41
    interopRequire
    0.40
    okin
    0.40
    ngrx
    0.40
    typeDefinition
    0.39
    Оп
    0.39
     kiss
    0.39
    enterOuterAlt
    0.38
    Activations Density 0.004%

    No Known Activations