© 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 - 131k
    5. 20-GEMMASCOPE-RES-131K
    6. 73581
    Prev
    Next
    INDEX
    Explanations

    symbols or formatting elements that denote emphasis or special significance

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

    * followed by a letter

    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
     ainfi
    -0.60
    pecabe
    -0.57
     anún
    -0.56
     romántico
    -0.55
     geométrico
    -0.55
     viņ
    -0.52
     valentín
    -0.51
     instituição
    -0.51
     tilbage
    -0.50
     farmacé
    -0.50
    POSITIVE LOGITS
     SOP
    0.63
     Dider
    0.59
    SOP
    0.57
    soni
    0.57
    sibirsk
    0.54
    asten
    0.54
     Sapporo
    0.54
    ipts
    0.54
    zeera
    0.54
    leine
    0.54
    Activations Density 0.028%

    No Known Activations