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. 74095
    Prev
    Next
    INDEX
    Explanations

    technical terms and factors related to engineering and material science

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

    prevention, protection, and effectiveness

    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
    はじめに
    -0.45
    asteroide
    -0.44
     zebrafish
    -0.41
    StoreMessageInfo
    -0.41
     المعيارى
    -0.40
    chromedriver
    -0.40
     mitään
    -0.40
    ècle
    -0.40
     citoy
    -0.39
     citoyens
    -0.39
    POSITIVE LOGITS
    の効果
    0.52
     prevention
    0.52
     intervention
    0.51
     increase
    0.50
     treatment
    0.48
     effectiveness
    0.47
     protection
    0.46
    prevention
    0.46
     effective
    0.45
     protect
    0.45
    Activations Density 0.103%

    No Known Activations