© 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
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 11863
    Prev
    Next
    INDEX
    Explanations

    technical terminology related to vehicle maintenance and tire components

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_20/width_16k/average_l0_68
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    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
    "])){
    -0.37
     affaires
    -0.37
     gorro
    -0.36
     doğal
    -0.36
     ehre
    -0.35
     gestes
    -0.35
     levier
    -0.35
     mémor
    -0.34
    ActionListener
    -0.34
    Personen
    -0.34
    POSITIVE LOGITS
     tyre
    0.90
     tyres
    0.80
     AssemblyCulture
    0.75
     Tire
    0.74
     tires
    0.69
     tire
    0.68
     Tyres
    0.67
     Tires
    0.67
    OGND
    0.65
    RTLI
    0.65
    Activations Density 0.071%

    No Known Activations