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. 2907
    Prev
    Next
    INDEX
    Explanations

    phrases related to instructions for preparation and conditions for use

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

    time, measure, or cost

    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
    providedIn
    -0.48
    GraphicsUnit
    -0.47
    Biografie
    -0.47
    termilk
    -0.45
    setHorizontal
    -0.43
    ----</
    -0.42
    itsub
    -0.42
     Nieuwe
    -0.42
    windowFixed
    -0.42
    dicionado
    -0.42
    POSITIVE LOGITS
     individuel
    0.35
    Hochspringen
    0.34
    rinos
    0.34
    addEventListener
    0.33
    日閲覧
    0.33
     informée
    0.33
    hafter
    0.33
     disambiguazione
    0.33
    イナー
    0.32
    Revenir
    0.32
    Activations Density 0.009%

    No Known Activations