© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 3916
    Prev
    Next
    INDEX
    Explanations

    expressions of greeting or attention, particularly the word "hey."

    oai_token-act-pair · gpt-4o-mini
    Triggered by @bot

    hey, bro

    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
    Negative Logits
    ArgsConstructor
    -0.83
    はじめに
    -0.81
     propOrder
    -0.80
     typelib
    -0.78
    matorium
    -0.77
    randomUUID
    -0.76
     FetchType
    -0.75
     للاسماء
    -0.73
     Савезне
    -0.73
    rungsseite
    -0.72
    POSITIVE LOGITS
     hey
    1.04
    Hey
    1.01
     Hey
    0.94
    hey
    0.91
     HEY
    0.85
    HEY
    0.77
     guys
    0.69
     hej
    0.62
    TagHelper
    0.61
     presto
    0.61
    Activations Density 0.004%

    No Known Activations