© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    5. Gemma-2-9B-IT
    7. Residual Stream - 131k
    9. 20-GEMMASCOPE-RES-131K
    11. 58254
    Prev
    Next
    INDEX
    Explanations

    instances of the word "coat" and its variations in the context of coatings or layers

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

    coat, coated, coating

    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
    displaymath
    -0.64
    URLException
    -0.55
    NSUserDefaults
    -0.54
     Jamestown
    -0.54
     TMPro
    -0.52
    SBATCH
    -0.52
    ElementException
    -0.51
    mediately
    -0.50
     MainAxisSize
    -0.50
    ynchronously
    -0.50
    POSITIVE LOGITS
    coat
    0.94
    Coat
    0.91
     coat
    0.87
     Coating
    0.86
     Coat
    0.85
     coating
    0.79
     Coats
    0.74
    coating
    0.73
     Coated
    0.70
     coats
    0.68
    Activations Density 0.007%

    No Known Activations