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. 15013
    Prev
    Next
    INDEX
    Explanations

    technical terms related to electrical components and circuitry

    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
     algorithms
    -0.37
     Amb
    -0.35
     algorithm
    -0.35
     Algorithms
    -0.35
     manip
    -0.34
     Force
    -0.34
     diesel
    -0.33
     algorit
    -0.33
    getStatusCode
    -0.33
     Navi
    -0.33
    POSITIVE LOGITS
     tubes
    0.84
     tube
    0.84
    tubes
    0.77
     Tubes
    0.76
    Tube
    0.74
     Tube
    0.74
    tube
    0.73
     TUBE
    0.68
     tubos
    0.66
    vacuum
    0.59
    Activations Density 0.607%

    No Known Activations