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. 7662
    Prev
    Next
    INDEX
    Explanations

    references to quantities and locations within a historical or geographical context

    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
     kasarigan
    -0.61
     suburban
    -0.51
     suburbs
    -0.49
     suburb
    -0.47
     Numerade
    -0.46
    issauga
    -0.45
    FetchType
    -0.45
     nonUne
    -0.44
     Minneapolis
    -0.43
     Indianapolis
    -0.43
    POSITIVE LOGITS
     Historic
    0.58
    Historic
    0.52
     courthouse
    0.50
     historic
    0.50
     InputDecoration
    0.48
    stdc
    0.47
    Rotary
    0.44
     Historical
    0.44
     Rotary
    0.43
     Depot
    0.43
    Activations Density 0.318%

    No Known Activations