© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    5. Gemma-2-2B
    7. Residual Stream - 65k
    9. 20-GEMMASCOPE-RES-65K
    11. 3506
    Prev
    Next
    INDEX
    Explanations

    references to "crazy" or deviant behavior

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

    cra followed by common suffixes

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 20-gemmascope-res-65k
    Configuration
    google/gemma-scope-2b-pt-res/layer_20/width_65k/average_l0_114
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    65,536
    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
     kaarangay
    -0.80
    expandindo
    -0.73
    +#+#
    -0.70
    
    -0.69
     resourceCulture
    -0.66
    WEBPACK
    -0.66
    تقاوى
    -0.65
     Boron
    -0.65
    PyErr
    -0.65
     GetEnumerator
    -0.63
    POSITIVE LOGITS
    cra
    0.89
    Cra
    0.87
     cra
    0.86
     Cra
    0.80
     cradle
    0.60
     sprüche
    0.57
     viewDidLoad
    0.54
     Craven
    0.53
    opropyl
    0.52
    moiselle
    0.51
    Activations Density 0.017%

    No Known Activations