Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B-IT
    4. Residual Stream - 131k
    5. 20-GEMMASCOPE-RES-131K
    6. 85252
    Prev
    Next
    INDEX
    Explanations

    words related to crushing and being crushed

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

    crush and variations

    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
     للمعارف
    -0.77
    ⓧ
    -0.68
     getItemId
    -0.63
    ſicht
    -0.60
     intptr
    -0.56
    Portail
    -0.56
    VIAF
    -0.56
    hyrchwyd
    -0.55
    elementType
    -0.54
    argout
    -0.54
    POSITIVE LOGITS
    crush
    1.06
     crush
    1.05
     Crush
    1.05
    Crush
    1.02
     crushing
    0.87
     crushed
    0.81
     crusher
    0.75
     crushes
    0.73
    Crus
    0.66
    shat
    0.65
    Activations Density 0.003%

    No Known Activations