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
    4. Residual Stream - 16k
    5. 38-GEMMASCOPE-RES-16K
    6. 8689
    Prev
    Next
    INDEX
    Explanations

    instances of the word "in" across various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 38-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_38/width_16k/average_l0_128
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.38.hook_resid_post
    Hook Layer
    38
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    大
    -0.54
    地上
    -0.50
    线程
    -0.47
    地面
    -0.47
    地图
    -0.46
    大脑
    -0.46
    线性
    -0.45
    地铁
    -0.45
     внутрен
    -0.45
    地
    -0.44
    POSITIVE LOGITS
     accordance
    1.27
     addition
    1.23
     terms
    1.21
     spite
    1.08
     conjunction
    1.07
     relation
    1.04
     lieu
    0.95
     regards
    0.95
     order
    0.86
     vitro
    0.83
    Activations Density 1.189%

    No Known Activations