© 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 - 16k
    9. 22-GEMMASCOPE-RES-16K
    11. 601
    Prev
    Next
    INDEX
    Explanations

    phrases and actions related to conflict or confrontation

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

    Actions involving physical harm or close calls

    np_acts-logits-general · gemini-2.0-flash

    physical actions and reactions

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 22-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_22/width_16k/average_l0_72
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.22.hook_resid_post
    Hook Layer
    22
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    DeleteBehavior
    -0.77
     defStyle
    -0.62
     المعيارى
    -0.61
     ProtoMessage
    -0.60
    hyrchwyd
    -0.60
     يتيمه
    -0.57
    RegressionTest
    -0.57
     rapporte
    -0.56
    rrer
    -0.56
    jooq
    -0.55
    POSITIVE LOGITS
     dodged
    0.78
     dodge
    0.76
     swung
    0.66
     react
    0.66
     deflect
    0.64
     lunge
    0.62
     parry
    0.62
     reacted
    0.60
     swing
    0.60
     deflected
    0.60
    Activations Density 0.279%

    No Known Activations