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-2-2B
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 13794
    Prev
    Next
    INDEX
    Explanations

    phrases related to safety and security measures

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

    before "fence," "wall," or "rail"

    np_acts-logits-general · gemini-2.0-flash

    fences, walls, and barriers

    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-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    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
    JsonKey
    -0.59
    amaño
    -0.58
    ActionCreators
    -0.57
    atshirt
    -0.55
    requestCode
    -0.54
    onViewCreated
    -0.53
    Источники
    -0.51
    saraba
    -0.51
    EDEFAULT
    -0.51
     exponents
    -0.51
    POSITIVE LOGITS
     fence
    1.76
     fences
    1.60
     Fence
    1.47
     fencing
    1.43
     wall
    1.42
     barrier
    1.41
    fence
    1.38
     walls
    1.35
     barriers
    1.34
    Fence
    1.32
    Activations Density 0.141%

    No Known Activations